ICM Satellite Conference on

Probability and Stochastic Processes

Indian Statistical Institute, Bangalore
August 13-17, 2010.

Scientific Program, Abstracts, and Directory of Participants

Sponsors
 ICM Satellite Conference on
Probability and Stochastic Processes
Indian Statistical Institute, Bangalore
August 13-17, 2010.

Scientific Program, Abstracts, and Directory of Participants

Sponsors
Scientific Program Committee
David Aldous Univ. of California, Berkeley
Vivek S. Borkar TIFR, Mumbai
Mu Fa Chen Beijing Normal Univ., Beijing
Alice Guionnet ENS de Lyon, Lyon
Takashi Kumagai Kyoto University, Kyoto
Edwin A. Perkins Univ. of British Columbia, Vancouver
Rahul Roy Indian Statistical Institute, Delhi
Marta Sanz-Sole Univ. de Barcelona, Barcelona
Maria Eulalia Vares CBPF, Rio de Janeiro
Ofer Zeitouni Weizmann Inst. of Science, Rehovot

Local organising committee

Siva Athreya Indian Statistical Institute, Bangalore
Abhay G. Bhatt Indian Statistical Institute, Delhi
Vivek S. Borkar Tata Institute of Fundamental Research, Mumbai
Arup Bose Indian Statistical Institute, Kolkata
Mrinal K. Ghosh Indian Institute of Science, Bangalore
Srikanth Iyer Indian Institute of Science, Bangalore
Rajeeva L. Karandikar Chennai Mathematical Institute, Chennai
Krishanu Maulik Indian Statistical Institute, Kolkata
B. Rajeev Indian Statistical Institute, Bangalore
S. Ramasubramanian Indian Statistical Institute, Bangalore
Anish Sarkar Indian Statistical Institute, Delhi
Rajesh Sundaresan Indian Institute of Science, Bangalore

Jointly Organised by
 Scientific Program Overview

Friday, August 13, 2010

08:00 - 09:00 Registration / Breakfast

Alain-Sol Sznitman, ETH, Zurich

09:00 - 10:00 Random Interlacements
Plenary Session Chair : Antar Bandhyopadyay, Indian Statistical Institute, New Delhi
09:00 -- 11:00
Auditorium (main) Andrea Montanari, Stanford University
10:00 - 11:00 Message passing algorithms, random convex problems, and the risk of the LASSO
Chair : B.V. Rao, Chennai Mathematical Institute

11:00 -- 11:30 Coffee Break

Organiser & Chair : Erwin Bolthausen, University of Zurich

Noam Berger, Hebrew University of Jerusalem

11:30 - 12:05 Unsatisfied edges in 2D Spin Glass ground states do not percolate
Invited Session
Random Media
11:30 -- 13:15 Vladas Sidoravicius, CWI/IMPA
12:05 -- 12:40 Random walks and random growth in dynamic random environment
Auditorium (main)

Gerard BEN AROUS, Courant Institute, New York University

12:40 -13:15 Complexity of Spin Glass energy landscapes and Random Matrices

Organiser & Chair : Arturo Kohatsu-Higa, Osaka University

Giovanni Peccati, Luxembourg University

11:30 -- 12:05
Invited Session Universal Gaussian Fluctuations on Wiener Chaos
Malliavin
Calculus
11:30 -- 13:15 12:05 -- 12:40 Jonathan Mattingly, Duke University
SSIU seminar hall

Arturo Kohatsu Higa, Osaka University

12:40 -- 13:15 A Malliavin Calculus method to study SDE's with irregular drifts

Chair : Leonid Mytnik, Technion - Israel Institute of Technology

Alberto Algergo Pinto, University of Porto

Invited Session 11:30 -- 12:05 Hotelling model with uncertainty on the production cost and Networks
Games and
Control
11:30 -- 13:15 K.S. Mallikarjuna Rao, Indian Institute of Technology Bombay
Auditorium 12:05 -- 12:40 Nash Equilibrium in Stochastic Games and its Approximation
(seminar hall)

Anup Biswas, TIFR Centre for Applicable Mathematics

12:40 -- 13:15 Exit Time and Invariant Measure Asymptotics for Small Noise Constrained Diffusions

13:15 -- 14:30 Lunch Break

13:15 -- 14:30 Lunch Break

Chair : Mokshay Madiman, Yale University

Organiser : Sourav Chatterjee, University of California, Berkeley

Kavita Ramanan, Brown University

14:30 -- 15:05 Concentration Inequalities for Dependent Random Variables
Invited Session
Concentration
Inequalities Andrea Montanari, Stanford University
14:30 -- 16:15 15:05 -- 15:40 Some concentration questions in random constraint satisfaction problems
Auditorium (main)

Pierre Del Moral, INRIA

15:40 -- 16:15 Concentration Inequalities for Mean Field Particle Models

Organiser & Chair : Martin Hairer, University of Warwick

Jeremy Quastel, University of Toronto
14:30 -- 15:05 Exact distribution function for the free energy of the continuum directed random
Invited Session polymer and KPZ in one dimension
SPDE
14:30 -- 16:15 Leonid Mytnik, Technion - Israel Institute of Technology
SSIU seminar hall 15:05 -- 15:40 Infinite rate mutually catalytic branching model

Jan Maas, Universität Bonn

15:40 -- 16:15 Approximations to stochastic Burgers equations

Chair : Ayalvadi Ganesh, University of Bristol

Anish Sarkar, Indian Statistical Institute

14:30 -- 15:05 Random geometric graphs with densitites having a zero
Invited Session
Random Debleena Thacker, Indian Statistical Institute, Delhi
Geometric 15:05 -- 15:40 Non-Uniform Random Geometric Graphs with Location Dependent Radii
Graphs
14:30 -- 16:50
Auditorium Srikanth K Iyer, Indian Institute of Science Bangalore
(seminar hall) 15:40 -- 16:15
Extended Random Signal-to-Interference-and-Noise-Ratio Graphs with Fading

Yogeshwaran Dhandapani, Ecole Normale Superieure

16:15 -- 16:50 Percolation and Connectivity in AB Random Geometric Graphs.

Chair : Gabor Lugosi, ICREA and Pompeu Fabra University

Koushik Saha, Bidhannagar College, Kolkata

14:30 -- 15:05 Product of Exponentials and Spectral Radius of Random k Circulants

Invited Session Rasbagh Vasudeva, University of Mysore

Heavy Tails and 15:05 -- 15:40 Laws of the Iterated logarithm for trimmed sums under domains of partial attraction
Related Topics
14:30 -- 16:50
Auditorium (smu) Sreenivasan Ravi, University of Mysore
15:40 -- 16:15
On some tail bahaviours of distributions

Arunangshu Biswas, Presidency College

16:15 -- 16:50 Process convergence of self normalized sums of i.i.d. random variables coming from
domain of attraction of stable distributions

16:15 -- 17:00 Coffee Break

 Saturday, August 14, 2010
08:00 - 09:00 Registration / Breakfast

Maury Bramson, University of Minnesota

Stability and tail behavior for randomized load balancing schemes
Plenary Session Chair : B. Rajeev, Indian Statistical Institute, Bangalore
09:00 -- 11:00
Auditorium (main) Gabor Lugosi, ICREA and Pompeu Fabra University
Combinatorial testing problems
Chair : Srikanth Iyer, Indian Institute of Science

11:00 -- 11:30 Coffee Break

Organiser & Chair : Gopal Basak, Indian Statistical Institute, Kolkata

Hoi Ying Wong, The Chinese University of Hong Kong

11:30 -- 12:05 Mean-variance portfolio and asset-liability management when assets are co-integrated
Invited Session
Mathematical
Finance Freddy Delbaen, ETH, Zurich
11:30 -- 13:15 12:05 -- 12:40
BSDE with unbounded terminal value
SSIU seminar hall

K. Suresh Kumar, Indian Institute of Technology Bombay

12:40 --13:15 Stock models corresponding to modification in Black-Scholes formula

Chair : Manjunath Krishnapur, Indian Institute of Science

Organiser : Wendelin Werner, Universite Paris-Sud

Hugo Duminil-Copin, University of Geneva

11:30 -- 12:05 Self-avoiding walk on the hexagonal lattice
Invited Session
SLE
11:30 -- 13:15
Auditorium (main) Antti Kemppainen, University of Helsinki
12:05 -- 12:40 Random curves, scaling limits and Loewner evolutions

Pierre Nolin, Courant Institute, New York University

12:40 -- 13:15 Connection probabilities and RSW-type bounds for the two-dimensional FK Ising model

Chair : Suresh Kumar, Indian Institute of Technology Bombay

Evelina Shamarova, University of Porto

Invited Session 11:30 -- 12:05 Solutions of the Navier-Stokes and Burgers equations via forward-backward SDEs
Stochastic
Differential
Equations Lluís Quer-Sardanyons, Universitat Autònoma de Barcelona
11:30 -- 13:15 12:05 -- 12:40 Weak convergence for the stochastic heat equation driven by Gaussian white noise
Auditorium
(seminar hall)
Noèlia Viles, Universitat Autònoma de Barcelona
12:40 -- 13:15 Continuity in the Hurst parameter of the law of the Russo-Vallois symmetric integral

13:15 -- 14:30 Lunch Break

13:15 -- 14:30 Lunch Break

Organiser & Chair : Kavita Ramanan, Brown University

Shankar Bhamidi, University of North Carolina, Chapel Hill

14:30 -- 15:05 Flows, first passage percolation and random disorder in networks
Invited Session
Stochastic
Networks Lea Popovic, Concordia University
14:30 -- 16:15 15:05 -- 15:40 Diffusion approximations for multiscale reaction networks
SSIU seminar hall

Ayalvadi Ganesh, University of Bristol

15:40 -- 16:15 Decentralised load balancing in closed and open networks

Chair : Maury Bramson, University of Minnesota

Antar Bandyopadhyay, Indian Statistical Institute, Delhi

14:30 -- 15:05 On the expected total number of infections for virus spread on a finite network

Invited Session Partha Sarathi Dey, University of California, Berkeley

Interacting 15:05 -- 15:40 Stein's method for concentration inequalities - extension and applications
Particle Systems
14:30 -- 16:50
Auditorium (main) Raazesh Sainudiin, University of Canterbury
15:40 -- 16:15 Exact transition probabilities of the number of ancestral sample lineages in the
discrete ancestral recombination graph

Rahul Roy, Indian Statistical Institute, Delhi

16:15 -- 16:50 Learning from neighbours

Chair : S. R. S. Varadhan, Courant Institute, New York University

Ghurumuruhan Ganesan, Indian Statistical Institute, Delhi

14:30 -- 15:05 Two properties of the random partition of an integer
Invited Session
Combinatorial Krishanu Maulik, Indian Statistical Institute, Kolkata
Probability and 15:05 -- 15:40 Strong laws for balanced block triangular urns with irreducible diagonal blocks
Urn Models
14:30 -- 16:15
Auditorium Srinivasan Balaji, George Washington University
(seminar hall) 15:40 -- 16:15 Phases in the diffusion of gases via the Ehrenfest urn

16:15 -- 17:00 Coffee Break

Saturday, August 14, 2010

No Talks
 Monday, August 16, 2010
08:00 - 09:00 Registration / Breakfast

Erwin Bolthausen, University of Zurich

Non-ballistic random walks in random environments
Plenary Session Chair : Rahul Roy, Indian Statistical Institute, Delhi
09:00 -- 11:00
Auditorium (main) Alison Etheridge, University of Oxford
Modelling evolution in a spatial continuum
Chair : Krishanu Maulik, Indian Statistical Institute, Kolkata

11:00 -- 11:30 Coffee Break

Organiser & Chair : Samy Tindel, University of Nancy

Peter Friz, TU-Berlin and WIAS

11:30 - 12:05 A (rough) pathwise approach to SPDEs
Invited Session
Rough
Path Analysis Yuzuru Inahama, Nagoya University
11:30 -- 13:15 12:05 -- 12:40 Laplace-type asymptotics in the rough path theory
SSIU seminar hall

Samy Tindel, University of Nancy

12:40 -13:15 Rough paths and Skorohod stochastic calculus for Gaussian processes

Chair : Anish Sarkar, Indian Statistical Institute, Delhi

Organiser : Ken Alexander, University of Southern California

Hubert Lacoin, Universita di Roma Tre

11:30 -- 12:05 Directed polymers in random environment: Influence of disorder and the
Invited Session superdiffusivity phenomenon
Polymer Models
11:30 -- 13:15
Auditorium (main) Xia Chen, University of Tennessee
12:05 -- 12:40 Brownian motion in a renormalized Poissonian potential

Rongfeng Sun, National University of Singapore

12:40 -- 13:15 Random walk among moving Poissonian traps

Chair : Alison Etheridge, University of Oxford

Vivek Shripad Borkar, Tata Institute of Fundamental Research

Invited Session 11:30 -- 12:05
Small noise limits and Feller selection
Diffusions and
Related Models
11:30 -- 13:15 Krishnamurthi Ravishankar, SUNY-New Paltz
12:05 -- 12:40 Marking the Brownain web and applications
Auditorium
(seminar hall)
Siva Athreya, Indian Statistical Institute, Bangalore
12:40 -- 13:15
Brownian motion on R trees

13:15 -- 14:30 Lunch Break

13:15 -- 14:30 Lunch Break

Organiser & Chair : Arup Bose, Indian Statistical Institute, Kolkata

Steven J Miller, Williams College

14:30 -- 15:05 Eigenvalue statistics for Toeplitz ensembles
Invited Session
Random Matrices
14:30 -- 16:15 Manjunath Krishnapur, Indian Institute of Science
15:05 -- 15:40 Limit of characteristic polynomials of a random matrix
Auditorium (main)

Wenbo Li, University of Delaware

15:40 -- 16:15 Probabilities of all real zeros for random polynomials

Chair : Louis Chen, National University of Singapore

Tamal Banerjee, Indian Institute of Science

Invited Session 14:30 -- 15:05 Pricing of defaultable bonds in a Markov modulated market
Markov Models
and Related Topics
14:30 -- 16:15 Biswaroop Mookherjee, BASES The Nielsen Company
15:05 -- 15:40 Markov chains with random transition probability matrix
Auditorium
(seminar hall)
Subhamay Saha, Indian Institute of Science
15:40 -- 16:15 Generalised semi-Markov processes

Chair : Noam Berger, Hebrew University of Jerusalem

Rahul Jain, University of Southern California

14:30 -- 15:05 Solving Markov decision problems via simulations
Invited Session
Statistics and
Simulations Ananya Lahiri, Indian Institute of Technology, Kanpur
14:30 -- 16:15 15:05 -- 15:40 Estimating the parameters of multiple chirp signals
Auditorium (smu)

15:40 -- 16:15 Sandeep Juneja, Tata Institute of Fundamental Research

A large deviations perspective on efficiency of simulation estimators

Chair : Alain-Sol Sznitman, ETH, Zurich

Invited Session
Information Mokshay Madiman, Yale University
14:30 -- 15:05 Concentration of information content for log-concave probability measures
Theory and
Free Probability
14:30 -- 15:40
SSIU seminar hall Rajat Subhra Hazra, Indian Statistical Institute, Kolkata
15:05 -- 15:40 Subexponentiality of free regularly varying random variables

16:15 -- 17:00 Coffee Break

 Tuesday, August 17, 2010
08:00 - 09:00 Registration / Breakfast

Louis Chen, National University of Singapore

Normal approximation by Stein's method
Plenary Session Chair : Mrinal Ghosh, Indian Institute of Science
09:00 -- 11:00
Auditorium (main) Srinivasa Varadhan, Courant Institute, New York University
Central limit theorems for dependent random variables
Chair : Vivek Shripad Borkar, Tata Institute of Fundamental Research

11:00 -- 11:30 Coffee Break

Free
Friday, August 13, 2010

08:00 - 09:00 Registration

09:00 - 10:00 Plenary Session Venue: Auditorium (main)

(Chair: Antar Bandyopadhyay, Indian Statistical Institute, Delhi)

Alain-Sol Sznitman, ETH, Zurich

Random interlacements

Random interlacements are a Poissonian cloud of doubly infinite paths modulo time-shift, which yields a micro-
scopic model for the trace left at suitably selected time scales by random walks on large recurrent graphs, which
are locally transient. Random interlacements have recently been helpful in understanding, among other things,
the way in which random walk trajectories can create large disconnecting interfaces. We will review in this talk
some of the recent developments.
Keywords : random interlacements, random walks, percolation, disconnection
AMS Classification Numbers : 60K35, 60G50, 82C41

10:00 - 11:00 Plenary lecture: Andrea Montanari Venue: Auditorium (main)

(Chair: Bhamidi Visweswara Rao, Chennai Mathematical Institute)

Andrea Montanari, Stanford University

Message passing algorithms, random convex problems, and the risk of the LASSO

The problem of estimating a high-dimensional vector from a set of linear observations arises in a number of
engineering disciplines. It becomes particularly challenging when the underlying signal has some non-linear
structure that needs to be exploited. I will present a new class of iterative algorithms inspired by probabilistic
graphical models and statistical mechanics ideas, that appear to be asymptotically optimal in specific contexts.
The analysis of these algorithms allows to prove remarkably sharp results on the asymptotic behavior of some
families of random convex problems. I will in particular discuss the mean square error for LASSO estimation in
the context of compressed sensing problems. [Based on joint work with David L. Donoho and Arian Maleki, and
with Mohsen Bayati and Jose Bento.]
Keywords : message passing, convex geometry, LASSO, compressed sensing
AMS Classification Numbers : 60F99, 68W99, 62H12

11:00 - 11:30 Coffee Break

11:30 - 13:15 Invited Session: Random Media Venue: Auditorium (main)

(Chair: Erwin Bolthausen, University of Zurich)

11:30 Noam Berger, Hebrew University of Jerusalem

Unsatisfied edges in 2D spin glass ground states do not percolate

We define the Edwards-Anderson Ising Spin Glass model in the planar square lattice and survey some results
about it. We then define the notion of unsatisfied edges (in ground states of the above model) and prove that
they do not percolate.
This is joint work with Ran Tessler.
Keywords : spin glass, ground state, Edwards-Anderson model
AMS Classification Numbers : 60K35, 60K37

1
 12:05 Vladas Sidoravicius, CWI/IMPA
Random walks and random growth in dynamic random environment

I will focus on two questions: 1. Diffusion Limited Aggregation type growth model evolving in a dynamic random
environment, constituted by infinitely many particles, performing simple symmetric random walks independently
of each other, and sticking to the DLA-cluster, once they hit it. The number of particles in the system is preserved
during the evolution. The difference from the canonical Sander-Witten model is that in this case multiple particles
are simultaneously affecting shape of the cluster. I will show, that there is a phase transition in the initial density
of the environment particles, which separates linear and sub-linear growth regions, and, as a consequence,
limiting shapes of the aggregate.
2. Random walks in dynamic random environment. Consider a particle, call it W-particle, moving in a random
environment, where the jump probabilities of W-particle are affected by the environment. As in the model 1
above, the environment is constituted by the system of infinitely many particles performing simple symmetric
random walks, independently of each other, and once W-particle is on the empty vertex, it has drift in a direction
u, and if it shares a vertex with the environment particle it has opposite drift (−u). Again, there is a transition in
the behavior of W-particle, which depends on the density of environment particles.
What relates these two questions are techniques which we use analyze above models, and I will present some
technical aspects of multi-scale analysis, and regeneration time approach which are used in our work.
The talk is based on works in collaboration F. den Hollander, M. Hilario and H. Kesten.
Keywords : DLA, random walk in random environment, regeneration time
AMS Classification Numbers : 60K35, 60J25, 82C24

12:40 Gerard Ben Arous, Courant Institute, New York University

Complexity of spin glass energy landscapes and random matrices

This is joint work with C.A. Auffinger (Courant) and [Link] (ETH Zurich).
Random Matrix Theory and the Kac-Rice formula allow a very precise counting of the mean number of critical
points (of given index and given energy value) for spherical spin glasses hamiltonian. A detailed picture of the
shape of the energy landscape emerges, which is rather different for the pure p-spin case and for mixed cases.
I will dwell on recent progress for general cases.
Keywords : spin glasses, random matrices, complexity
AMS Classification Numbers : 15A52, 52C45, 60G15, 60K35

11:30 - 13:15 Invited Session: Malliavin Calculus Venue: SSIU seminar hall
(Chair: Arturo Kohatsu Higa, Osaka University)

11:30 Giovanni Peccati, Luxembourg University

Universal Gaussian fluctuations on Wiener chaos

We show how to compute explicit bounds in the normal and chi-square approximations of multilinear homoge-
nous sums (of arbitrary order) of general centered independent random variables with unit variance. Our tech-
niques combine an invariance principle by Mossel, O’Donnell and Oleszkiewicz with a refinement of some
recent results by Nourdin and Peccati, about the approximation of laws of random variables belonging to a
fixed (Gaussian) Wiener chaos. In particular, we will show that chaotic random variables enjoy the following
form of universality: (a) the normal and chi-square approximations of any homogenous sum can be completely
characterized and assessed by first switching to its Wiener chaos counterpart, and (b) the simple upper bounds
and convergence criteria available on the Wiener chaos extend almost verbatim to the class of homogeneous
sums. These results partially rely on the notion of “low influences” for functions defined on product spaces, and
provide a generalization of several central and non-central limit theorems for smooth functionals of Gaussian
fields. They also imply a further drastic simplification of the method of moments and cumulants – as applied
to the proof of probabilistic limit theorems – and yield substantial generalizations, new proofs and new insights

2
 into some classic findings by de Jong and Rotar’. Our tools involve the use of Malliavin calculus, and of both the
Stein’s method and the Lindeberg invariance principle for probabilistic approximations.

References
[1] I. Nourdin and G. Peccati (2009). Stein’s method on Wiener chaos. Probab. Theory Rel. Fields 145, no. 1,
75-118.
[2] I. Nourdin, G. Peccati and G. Reinert (2009). Invariance principles for homogeneous sums: universality of
Gaussian Wiener chaos. To appear in: Ann. Probab.
[3] D. Nualart and G. Peccati (2005). Central limit theorems for sequences of multiple stochastic integrals. Ann.
Probab. 33 (1), 177-193.
[4] E. Mossel, R. O’Donnell and K. Oleszkiewikz (2008). Noise stability of functions with low influences: variance
and optimality. To appear in: Ann. Math..
[5] G. Peccati and M.S. Taqqu (2010). Wiener Chaos: Moments, Cumulants and Diagrams. Springer Verlag.
Keywords : limit theorems, universality
AMS Classification Numbers : 60F05, 60F17, 60G15, 60H07

12:05 Jonathan Mattingly, Duke University

Malliavin Calculus to prove ergodic theorems for SPDEs

I will give a tour of a collections of techniques based on Malliavin calculus which have been developed to prove
ergodic theorems for degenerate diffusions on infinite dimensional spaces, such as the configuration space of
an Stochastic Partial Differential equation (SPDE). This will include a discussion of non-adapted estimates and
a partial integration by parts formula. The goal will be to prove an infinite dimensional version of Hormander’s
"Sum of Squares Theorem" for dissipative SPDEs. Such questions, in the finite dimensional setting were, the
original motivation for the development of Malliavin calculus. The talk will be based on joint work with Martin
Hairer, Etienne Paradox and Yuri Bakhtin.
Keywords : stochastic partial differential equation (SPDE), Malliavin calculus, ergodic theorem
AMS Classification Numbers : 37A25, 37A60, 37N10, 37L55, 60H15, 60H07, 35R60

12:40 Arturo Kohatsu Higa, Osaka University

A Malliavin calculus method to study SDE’s with irregular drifts

Until recently it was thought that Malliavin Calculus was a tool to be used with diffusions with smooth coefficients.
We present a general method which allows to use Malliavin Calculus for stochastic equations with irregular
drift. This method uses the Girsanov theorem combined with Ito-Taylor expansion in order to obtain regularity
properties for the density of a hypoelliptic random variable at a fixed time. We apply the methodology to the
case of the Lebesgue integral of a diffusion.
This is joint work with Akihiro Tanaka.
Keywords : Malliavin calculus, SDE with irregular drift
AMS Classification Numbers : 60F05, 60F10, 60F17

11:30 - 13:15 Contributed Session: Games and Control Venue: Auditorium (seminar hall)
(Chair: Leonid Mytnik, Technion - Israel Institute of Technology)

11:30 Alberto Algergo Pinto, University of Porto

Hotelling model with uncertainty on the production cost and networks

First, we find the ex-ante and ex-post prices of the Hotelling model with uncertainty on the production cost of
both firms. Secondly, we introduce a new Hotelling-type network game, where each firm i is represented by a

3
 node of degree ki , the number of firm i’s direct competitors (neighbors). We investigate price competition a la
Hotelling with complete and incomplete information about the network structure. The goal is to investigate the
effects of the network structure and of the uncertainty on firms’ prices and profits.

References
[1] A. A. Pinto, Game Theory and Duopoly Models. Interdisciplinary Applied Mathematics series, Springer-
Verlag (2010).
[2] Dynamics, Games and Science. Eds: M. Peixoto, A. A. Pinto and D. A. Rand. Proceedings in Mathematics
series, Springer-Verlag (2010).
[3] A. A. Pinto, D. A. Rand and F. Ferreira, Fine Structures of Hyperbolic Diffeomorphisms. Springer-Verlag
Monograph (2009).
Keywords : Hotelling model, networks, uncertainty
AMS Classification Numbers : 90B15, 91A43

12:05 K.S. Mallikarjuna Rao, Indian Institute of Technology Bombay

Nash equilibrium in stochastic games and its approximation

Using the fact that any two player discounted stochastic game with finite state and action spaces can be recast
as a non-convex constrained optimization problem, where each global minima corresponds to a stationary
Nash equilibrium, we present a sequential quadratic programming based algorithm that converges to a KKT
point. This KKT point is an -Nash equilibrium for some  > 0 and under some suitable conditions we show that
this KKT point corresponds to a stationary Nash equilibrium. The algorithm updates the Hessian matrix of the
Lagrangian function in a specific way. We illustrate various difficulties that can arise while computing stationary
Nash equilibrium of the stochastic game using a variant of pollution tax model. One interesting feature of this
model (in an instance) is that it admits a Nash equilibrium which is independent of the discount factor close to
1, an extension of Blackwell optimality in Markov decision processes.
Keywords : nonzero-sum stochastic game, discounted value, multiple Nash equilibria, -Nash equilibrium, non-
convex constrained optimization, sequential quadratic programming, Lagrange multipliers, MFCQ condition,
regular point, pollution tax model
AMS Classification Numbers : 91A15, 90C26, 91B76

12:40 Anup Biswas, TIFR Centre for Applicable Mathematics

Exit time and invariant measure asymptotics for small noise constrained diffusions

Constrained diffusions, with diffusion matrix scaled by small  > 0, in a convex polyhedral cone G ⊂ Rk , are
considered. Under suitable stability assumptions small noise asymptotic properties of invariant measures and
exit times from domains are studied. Let B ⊂ G be a bounded domain containing 0. Under conditions, an
“exponential leveling" property that says that, as  → 0, the moments of functionals of exit location from B ,
corresponding to distinct initial conditions coalesce asymptotically at an exponential rate, is established. Result
of this type in unconstrained settings was done by Martin V. Day. This result was obtained by using regularity
results from elliptic partial differential equations. But in the constrained settings we can not adopt the same
technique due to the irregular nature of boundary of the cone G. We use a probabilistic method based on
pseudo atom approach which bypasses the pde estimates. It is also shown that, with appropriate conditions,
difference of moments of a typical exit time functional with a sub-logarithmic growth, for distinct initial conditions
in suitable compact subsets of B , is asymptotically bounded. Furthermore, as initial conditions approach 0 at
rate 2 these moments are shown to asymptotically coalesce at an exponential rate. Result of these types are
totally new and not even known in unconstrained settings.
This work is done with Amarjit Budhiraja.
Keywords : large deviations, constrained diffusions, Skorohod problem, small noise asymptotics, exit time,
exponential leveling, split-chains, pseudo-atom, Lyapunov functions, invariant measures
AMS Classification Numbers : 60F10, 60J60, 60J25

4
13:15 – 14:30 Lunch

14:30 - 16:15 Invited Session: Concentration Inequalities Venue: Auditorium (main)

(Chair: Mokshay Madiman, Yale University)

14:30 Kavita Ramanan, Brown University

Concentration inequalities for dependent random variables

We will describe an approach that combines the martingale method with linear programming techniques to
derive concentration inequalities for dependent random variables. Applications to Markov chains and hidden
Markov chains will also be illustrated.
This is based on joint work with Leonid Kontorovich.
Keywords : concentration inequalities, martingale method, mixing coefficients, hidden Markov chains
AMS Classification Numbers : 60E15, 60J10, 60G42

15:05 Andrea Montanari, Stanford University

Some concentration questions in random constraint satisfaction problems

Let Z(F ) be the number of solutions of a random k -satisfiability formula F with n variables and clause density
α. It was proved by Ehud Friedgut that the property {Z(F ) > 0} undergoes a sharp threshold when the clause
density increases.
It is a common conjecture that, when Z(F ) > 0 with high probability, the normalized logarithm of the number of
solutions concentrates. Namely there should exists a non-random function α 7→ f (α) such that, for any δ > 0,
(1/n) log Z(F ) ∈ [f − δ, f + δ] with high probability.
We prove the conjecture under the stronger assumption that the probability that F is unsatisfiable is O(1/ log(n)1+a )
for a > 0. In particular, the assumption holds for all α < 1, which proves the above concentration claim in the
whole satisfiability regime of random 2-SAT. We also extend these results to a broad class of constraint satis-
faction problems.
Keywords : concentration of measure, constraint satisfaction problems, sharp thresholds
AMS Classification Numbers : 68R05, 60C05

15:40 Pierre Del Moral, INRIA

Concentration inequalities for mean field particle models

This talk is concerned with the fluctuations and the concentration properties of a general class of discrete gen-
eration and mean field particle interpretations of non linear measure valued processes. We combine an original
stochastic perturbation analysis with a concentration analysis for triangular arrays of conditionally independent
random sequences, which may be of independent interest. Under some additional stability properties of the
limiting measure valued processes, uniform concentration roperties with respect to the time parameter are also
derived. The concentration inequalities presented here generalize the classical Hoeffding, Bernstein and Ben-
nett inequalities for independent random sequences to interacting particle systems, yielding very new results
for this class of models. We illustrate these results in the context of McKean Vlasov type diffusion models,
McKean collision type models of gases, and of a class of Feynman-Kac distribution flows arising in stochastic
engineering sciences and in molecular chemistry.
This is joint work with Rio Emmanuel.

References
[1] Pierre Del Moral, Rio Emmanuel. Concentration inequalities for mean field particle models, HAL-INRIA
RR-6901 [21p], [submitted] (April 2009).
Keywords : concentration inequalities, mean field particle models, measure valued processes, Feynman-Kac
semigroups, McKean-Vlasov models

5
 AMS Classification Numbers : 65C35, 60B10

14:30 - 16:15 Invited Session: SPDE Venue: SSIU seminar hall

(Chair: Martin Hairer, University of Warwick)

14:30 Jeremy Quastel, University of Toronto

Exact distribution function for the free energy of the continuum directed random polymer and KPZ in
one dimension

We consider the solution of the stochastic heat equation ∂t Z = 12 ∂x2 Z − Z Ẇ with specially chosen initial
data, whose logarithm is the free energy of the continuum directed polymer, or the Hopf-Cole solution of the
Kardar-Parisi-Zhang equation.
Through a steepest descent analysis of the Tracy-Widom formulas for the asymmetric simple exclusion we
obtain explicit formulas for the one-dimensional marginal distributions.
This is applied to discrete directed polymers in a scaled, intermediate disorder regime, which interpolates be-
tween weak and strong disorder.
This is joint work with Tom Alberts, Gidi Amir, Ivan Corwin and Kostya Khanin.
Keywords : Kardar-Parisi-Zhang equation, stochastic heat equation, stochastic Burgers equation, random
growth, asymmetric exclusion process, anomalous fluctuations, directed polymers
AMS Classification Numbers : 82C22, 60H15

15:05 Leonid Mytnik, Technion - Israel Institute of Technology

Infinite rate mutually catalytic branching model

Consider the mutually catalytic branching process with finite branching rate γ . We show that as γ → ∞ this
process converges to a process that we call infinite rate mutually catalytic branching model. We study the
properties and in particular the long time behavior of the process.
This is a joint work with A. Klenke.
Keywords : mutually catalytic branching, stochastic partial differential equations, long time behavior
AMS Classification Numbers : 60G57, 60H15, 60J80

15:40 Jan Maas, Universität Bonn

Approximations to stochastic Burgers equations

We consider the 1D-stochastic Burgers equation driven by additive space-time white noise. Although this equa-
tion is well-posed, it turns out to be very unstable under approximations of the nonlinearity. In fact we shall
prove that various natural approximations converge to different limits. This phenomenon can be explained by
the spatial roughness of the solutions.
This is joint work with Martin Hairer.
Keywords : stochastic Burgers equations, finite differences, spatial regularity, concentration inequalities
AMS Classification Numbers : 60H15, 60H35, 35K55

14:30 - 16:50 Contributed Session: Venue: Auditorium (seminar hall)

Random Geometric Graphs
(Chair: Ayalvadi Ganesh, University of Bristol)

14:30 Anish Sarkar, Indian Statistical Institute, Delhi

Random geometric graphs with densities having a zero

6
 A random geometric graph is defined as follows : for a set of random points in some set A ⊂ Rd , fix r > 0
and connect a pair of points X , Y , provided that k X − Y k< r. The graph generated with the set of points
as the vertices and the random set of edges thus constructed, is called random geometric graph. Previous
studies have been conducted under the assumption that the random set of point originate as an iid sample
from a density function that is bounded away from 0 on A. We restrict our attention to densities on [0, 1] with
exactly one zero at 0. Further, we impose that the density is regularly varying at 0. Under this assumption, we
study asymptotic behaviour of connectivity distance of the random geometric graph. Our results show that the
connectivity distance behaves differently as compared to the case of densities that are bounded away from 0.
Keywords : random geometric graph, regularly varying function, inhomogeneous Poisson process
AMS Classification Numbers : 60C05, 05C80

15:05 Debleena Thacker, Indian Statistical Institute, Delhi

Nonuniform random geometric graphs with location dependent radii

We propose a distribution free approach to the study of random geometric graphs. The distribution of vertices
follows a Poisson point process with intensity function nf (·), where n ∈ N, and f is a probability density function
on Rd . A vertex located at x connects via directed edges to other vertices that are within a cut-off distance
rn (x). We prove strong law results for, (i) the critical cutoff function so that almost surely, the graph does not
contain any node with out-degree zero for sufficiently large n, (ii) the maximum and minimum vertex degrees.
We also provide a characterization of the cut-off function for which the number of nodes with out-degree zero
converges in distribution to a Poisson random variable. We illustrate this result for a class of densities with
compact support that have at most polynomial rates of decay to zero. Finally, we state a sufficient condition for
an enhanced version of the above graph to be almost surely connected eventually.
This is joint work with Srikanth Iyer.
Keywords : random geometric graph, uniform distribution, connectivity
AMS Classification Numbers : 60D05, 60G70, 05C05, 90C27

15:40 Srikanth K Iyer, Indian Institute of Science

Extended random signal-to-interference-and-noise-ratio graphs with fading

We study the asymptotic properties of a random geometric graph (SINR-F) on uniform points in which a directed
link exists between two nodes if the signal to interference-noise ratio is above a certain threshold. We first study
such a graph in the presence of fading effects alone (RGG-F). For this graph we prove an almost sure limit
for the critical power required to ensure that the graph does not possess isolated nodes and a criterion under
which the number of isolated nodes converges in distribution to a Poisson distribution. We derive a sufficient
condition under which the graph will be connected with high probability and derive almost sure bounds on the
maximum and minimum vertex degrees. We then prove an almost sure upper bound on the maximum received
interference. This allows us to choose an asymptotic spread parameter so as to bound the maximum received
interference. With this choice of spread parameters we can extend the results obtained for RGG-F to SINR-F.
Keywords : random geometric graphs, signal to interference noise ratio, connectivity, vertex degree
AMS Classification Numbers : 60D05, 60G70, 05C05, 90C27

16:15 Yogeshwaran Dhandapani, Ecole Normale Superieure

Percolation and connectivity in AB random geometric graphs

Given two independent Poisson point processes Φ(1) ,Φ(2) in Rd , the AB Poisson Boolean model is the graph
with points of Φ(1) as vertices and with edges between any pair of points for which the intersection of balls
of radius 2r centred at these points contains at least one point of Φ(2) . This is a generalization of the AB
percolation model on discrete lattices. We show the existence of percolation for all d ≥ 2 and derive bounds
for a critical intensity. We also provide a characterization for this critical intensity when d = 2. To study the
connectivity problem, we consider independent Poisson point processes of intensities n and cn in the unit cube.

7
 The AB random geometric graph is defined as above but with balls of radius r. We derive a weak law result for
the largest nearest neighbour distance and almost sure asymptotic bounds for the connectivity threshold.
This is joint work with Srikanth Iyer.
Keywords : random geometric graph, percolation, connectivity, wireless networks, secure communication
AMS Classification Numbers : 60D05, 60G70, 05C05, 90C27

14:30 - 16:50 Contributed Session: Venue: Auditorium (smu)

Heavy Tails and Related Topics
(Chair: Gabor Lugosi, ICREA and Pompeu Fabra University)

14:30 Koushik Saha, Bidhannagar College

Product of exponentials and spectral radius of random k circulants

Even though the distribution of the g fold product of independent and identi- cally distributed (i.i.d.) random
variables have been studied in the literature, there does not seem to be much work on its tail behaviour, when
they are in Gumbel domain of attraction. In this talk we first consider the tail behaviour when the random
variables are exponentially distributed.
Then we consider n × n random k -circulant matrices with n → ∞ and k = k(n) whose input sequence {al }l≥0
is i.i.d. with finite (2 + δ) moment. We study the asymptotic distribution of the spectral radius, when n = k g + 1.
We show that with appropriate scaling and centering, the limit distribution is Gumbel. We also identify the
centering and scaling constants explicitly. The proof uses appropriate normal approximation techniques and the
above tail behaviour.
This is joint work with Arup Bose and Rajat Subhra Hazra.
Keywords : tail of product, eigenvalues, k -circulant matrix, spectral radius, normal approximation
AMS Classification Numbers : 60B20, 60B10, 60F05, 62E20

15:05 Rasbagh Vasudeva, University of Mysore

Laws of the iterated logarithm for trimmed sums under domains of partial attraction

In this paper, we obtain the law of the iterated logarithm for the trimmed sums, when the random variables are
i.i.d. and in the domain of partial attraction of a semi stable law. In particular, when the random variables are
positive valued, we prove law of the iterated logarithm for trimmed sums with a delayed start and also establish
an associated renewal theorem.
Keywords : law of the iterated logarithm, trimmed sums, domain of partial attraction, semistable laws
AMS Classification Numbers : 60F15

15:40 Sreenivasan Ravi, University of Mysore

On some tail bahaviours of distributions

The talk is based on joint work with my former student Mrs. Praveena.
Consider distributions belonging to the max domain of attraction of the Frechet law under power normalization.
The proposed talk shall be on some von-Mises type sufficient conditions for a distribution to belong to the max
domain of attraction of the Frechet law under power normalization. Several examples illustrating the variety
of tail behaviours are included in the talk. The latter part of the talk shall be on distributions that are both
subexponential and in the max domain of attraction of the Frechet law under power normalization.
Keywords : Frechet law, power normalization, subexponential distributions
AMS Classification Numbers : 60G70, 60E05

8
 16:15 Arunangshu Biswas, Presidency College
Process convergence of self normalized sums of i.i.d. random variables coming from domain of
attraction of stable distributions

In this paper we show that the continuous version of the self normalised process

Sn (t) (nt − [nt])X[nt]+1

Yn,p (t) = + ,
Vn,p Vn,p

where
[nt] n
X X 1
Sn (t) = Xi , Vn,p = ( | Xi |p ) p ,
i=1 i=1

and Xi are i.i.d. random variables belonging to DA(α), has a non trivial distribution iff p = α = 2. The case for
2 > p > α and p ≤ α < 2 is systematically eliminated by showing that either of tightness or finite dimensional
convergence to a non-degenerate limiting distribution does not hold. This work is an ex- tension of the work by
Csörgö et al. who showed Donsker’s theorem for Yn,2 (·), i.e., for p = 2, holds iff α = 2 and identified the limiting
process as standard Brownian motion in sup norm.
This is joint work with Gopal Basak.
Keywords : self-normalised sums, invariance principles, domain of attraction, stable distributions
AMS Classification Numbers : 60F17, 60G52

16:15 – 17:00 Coffee Break

9
Saturday, August 14, 2010

08:00 - 09:00 Registration

09:00 - 10:00 Plenary lecture: Maury Bramson Venue: Auditorium (main)

(Chair: Rajeev Bhaskaran, Indian Statistical Institute, Bangalore )

Maury Bramson, University of Minnesota

Stability and tail behavior for randomized load balancing schemes

In this talk, we consider randomized load balancing schemes where an arriving job joins the shortest of d
randomly chosen queues from among a pool of n queues. Vvekenskaya, Dobrushin, and Karpelevich (1996)
considered the case with Poisson input and exponentially distributed service times, and derived an explicit
formula for the equilibrium distribution for fixed d as n → ∞. Since its tail decays doubly exponentially fast, this
distribution is useful in various applications.
Relatively little work has been done for general service times or input. For general service times, the behavior
of the service rule at each queue will now play a role in the behavior of the system. Even the question of under
which conditions the system is stable (i.e., its underlying Markov process is positive recurrent) for fixed n is no
longer obvious. One also wishes to understand the limiting tail behavior for such equilibria (provided they exist)
as n → ∞, as in the first paragraph.
Here, we discuss the stability of such systems, and extensions of these results that show the associated equilib-
ria are tight when restricted to a finite number of queues and n → ∞. Although the general question regarding
the limiting behavior of the equilibria as n → ∞ is open, certain important cases, such as the FIFO service rule,
can be analyzed. We also discuss a general ansatz that enables the analysis of a larger family of systems. This
material includes joint work with Y. Lu and B. Prabhakar.
Keywords : stochastic network, join the shortest queue, stability
AMS Classification Numbers : 60K25, 68M20, 90B15

10:00 - 11:00 Plenary lecture: Gabor Lugosi Venue: Auditorium (main)

(Chair: Srikanth K Iyer, Indian Institute of Science )

Gabor Lugosi, ICREA and Pompeu Fabra University

Combinatorial testing problems

We study a class of hypothesis testing problems in which, upon observing the realization of an n-dimensional
Gaussian vector, one has to decide whether the vector was drawn from a standard normal distribution or,
alternatively, whether there is a subset of the components belonging to a certain given class of sets whose
elements have been “contaminated,” that is, have a mean different from zero. We establish some general
conditions under which testing is possible and others under which testing is hopeless with a small risk. The
combinatorial and geometric structure of the class of sets is shown to play a crucial role. The bounds are
illustrated on various examples.
Keywords : hypothesis testing, multiple hypotheses, Gaussian processes
AMS Classification Numbers : 62F03, 62F05

11:00 - 11:30 Coffee Break

11:30 - 13:15 Invited Session: Mathematical Finance Venue: SSIU seminar hall
(Chair: Gopal K Basak, Indian Statistical Institute, Kolkata)

11:30 Hoi Ying Wong, The Chinese University of Hong Kong

Mean-variance portfolio and asset-liability management when assets are co-integrated

10
 Co-integration of major financial markets around the globe is well evidenced with strong empirical support.
Nowadays, it becomes indispensable for financial companies, which participate in the international financial
markets, to manage their assets and liabilities by taking into account the concept of co-integration. This paper
considers the continuous-time mean-variance (MV) portfolio and asset-liability management (ALM) problems in
an incomplete financial market with co-integrated assets. Therefore, the number of trading assets can be less
than the number of Brownian motions spanning the market. By the celebrated Granger (1981) representation
theorem, co-integrated time series should follow the error-correction model. We assume that the co-integration
market follows the diffusion limit of the error-correction model derived by Duan and Pliska (2004). Using the
Markowitz (1952) mean-variance portfolio criterion, we consider financial companies’ problems of minimizing
the variance of terminal wealth given an expected terminal wealth for the case of portfolio management and
minimizing the variance of surplus given an expected surplus for the case of ALM. These two problems are
collectively formulated as an MV portfolio problem with random parameters. The particular structure of co-
integration enables us to completely solve the portfolio and ALM problems in the sense that solutions of the
continuous-time portfolio policy and the efficient frontier are obtained as explicit and closed-form formulas. The
key is the recognition of an affine form in the solution of a system of backward stochastic differential equations.
This is joint work with Mei Choi Chiu.
Keywords : co-integration, mean-variance portfolio theory, asset-liability management, random parameters
AMS Classification Numbers : 91G10, 91G80

12:05 Freddy Delbaen, ETH, Zurich

BSDE with unbounded terminal value

Using some à priori estimates, we will show that a BSDE with convex square bounded driver and terminal value
having good exponential moments, has a unique solution in the same class.
This is joint work with Hu and Richou both from Université de Rennes.
Keywords : BSDE, unbounded terminal values, monetary utility function
AMS Classification Numbers : 91B16, 91B06, 91G80, 35H30, 49L25, 60H30

12:40 K. Suresh Kumar, Indian Institute of Technology Bombay

Stock models corresponding to modification in Black-Scholes formula

In this talk we study the effects of perturbation of the Black-Scholes formula in the stock price dynamics. As a
result, we get a general functional from for the volatility of the stock.
Keywords : Black-Scholes formula, local voltility models, Feynmann-Kac representation
AMS Classification Numbers : 91B70

11:30 - 13:15 Invited Session: SLE Venue: Auditorium (main)

(Chair: Manjunath Krishnapur, Indian Institute of Science)

11:30 Hugo Duminil-Copin, University of Geneva

Self-avoiding walk on the hexagonal lattice

We will prove a conjecture made by B. Nienhuis regarding the connective constant of the hexagonal lattice. More
precisely, we will show that the number an of self-avoiding walks of length n (starting at the origin) satisfies:
q
1 √
lim ann = 2+ 2.
n→∞

The proof uses a parafermionic observable for the self avoiding walk, which satisfies a half of the discrete
Cauchy-Riemann relations. Establishing the other half of the relations (which conjecturally holds in the scaling
limit) would also imply convergence of the self- avoiding walk to SLE(8/3). This is a joint work with S. Smirnov.

11
 Keywords : self-avoiding walks, conformal invariance, honeycomb Lattice
AMS Classification Numbers : 60C05, 82C41

12:05 Antti Kemppainen, University of Helsinki

Random curves, scaling limits and Loewner evolutions

Random curves arise naturally as interfaces in the 2D statistical physics and its lattice models. A general
strategy to prove the convergence of a random discrete curve, as the lattice mesh goes to zero, is first to
establish precompactness of the law giving the existence of subsequential scaling limits and then to prove the
uniqueness. In this talk, I will introduce a sufficient condition that guarantees the precompactness and also
that the limits are Loewner evolutions, i.e. they correspond to continuous Loewner driving processes. This
framework of estimates is applicable in almost all proofs aiming to establish that an interface converges to a
Schramm-Loewner evolution (SLE). In principle, it can be applied beyond SLE.
This is joint work with Stanislav Smirnov, Université de Genève.
Keywords : random curve, scaling limit, Schramm-Loewner evolution, weak convergence, arms exponent
AMS Classification Numbers : 60J67, 82B20

12:40 Pierre Nolin, Courant Institute, New York University

Connection probabilities and RSW-type bounds for the two-dimensional FK Ising model

For two-dimensional independent percolation, Russo-Seymour-Welsh (RSW) bounds on crossing probabilities

are an important a-priori indication of scale invariance, and they turned out to be instrumental to describe the
phase transition. They are in particular a key tool to derive the so-called scaling relations, that link the critical
exponents associated with the main macroscopic functions.

In this talk, we prove RSW-type uniform bounds on crossing probabilities for the FK Ising model at criticality,
independent of the boundary conditions. A central tool in our proof is Smirnov’s fermionic observable for the
FK Ising model, that makes some harmonicity appear on the discrete level, providing precise estimates on
boundary connection probabilities. We also prove several related results – including some new ones – among
which the fact that there is no magnetization at criticality, tightness properties for the interfaces, and the value
of the half-plane one-arm exponent.

This is joint work with H. Duminil-Copin and C. Hongler.

Keywords : Ising model, FK percolation, crossing probability, Russo-Seymour-Welsh
AMS Classification Numbers : 60K35, 82B20, 82B27

11:30 - 13:15 Contributed Session: Venue: Auditorium (seminar hall)

Stochastic Differential Equations
(Chair: K. Suresh Kumar, Indian Institute of Technology Bombay)

11:30 Evelina Shamarova, University of Porto

Solutions of the Navier-Stokes and Burgers equations via forward-backward SDEs

We establish a connection between the strong solution to the spatially periodic Navier-Stokes equations and a
solution to a system of forward-backward stochastic differential equations (FBSDEs) on the group of volume-
preserving diffeomorphisms of a flat torus. Assuming the existence of a solution to the Navier-Stokes equations
with the initial data in the Sobolev space H s for sufficiently large s, we construct a solution of the associated
system of FBSDEs. Conversely, if we assume that a solution of the system of FBSDEs exists, then the solution
of the Navier-Stokes equations can be obtained from the solution of the FBSDEs. In fact, the constructed FB-
SDEs on the group of volume-preserving diffeomorphisms can be regarded as an alternative characterization
to the Navier-Stokes equations for studying the properties of the latter. On the other hand, we describe a prob-

12
 abilistic construction of H s -regular solutions to the spatially periodic Burgers equation by proving the existence
and uniqueness theorem for the associated forward-backward stochastic system. This work is joint with A. B.
Cruzeiro.

References
[1] A. B. Cruzeiro and E. Shamarova. Navier-Stokes equations and forward-backward SDEs on the group of
diffeomorphisms of a torus, Stochastic Processes and their Applications, 119 (2009), 4034-4060.
[2] A. B. Cruzeiro and E. Shamarova. On a forward-backward stochastic system associated to the Burgers
equation, Proceedings of the Workshop on Stochastic Analysis & Finance, 2009.
Keywords : Navier-Stokes equations, forward-backward SDEs, Burgers equation, diffeomorphism group
AMS Classification Numbers : 65H10, 35Q30, 35Q53

12:05 Lluís Quer-Sardanyons, Universitat Autònoma de Barcelona

Weak convergence for the stochastic heat equation driven by Gaussian white noise

In many applications, even though the randomness acting on a certain system is neither white nor Gaussian,
one usually justifies somehow that the noisy inputs can be approximated by a Gaussian white noise (see e.g.
[2]). We illustrate this fact by considering the one-dimensional stochastic heat equation

∂Un ∂ 2 Un
(t, x) − 2 (t, x) = b(Un (t, x)) + θn (t, x), (t, x) ∈ [0, T ] × [0, 1] (1)
∂t ∂ x
with some initial condition and Dirichlet boundary conditions, where we assume that the family of processes
θn have square-integrable paths and their integral processes converge to the Brownian sheet. We establish
sufficient conditions on θn so that the mild solution of (1) converges in law, in the space of continuous func-
tions C([0, T ] × [0, 1]), to the mild solution of an analogous equation where θn (t, x) is formally replaced by
the space-time white noise Ẇ (t, x). For this, we first make use of a suitable continuous functional of the
stochastic convolution term in order to reduce the proof to the linear version of (1). Secondly, we prove that
the corresponding family of laws is tight and we identify the limit law by showing the convergence of the finite
dimensional distributions.
Moreover, we consider two particular families of noises θn to which our result applies. The first one involves
a Poisson process in the plane and has
R t been motivated by a one-dimensional result of Stroock ([1]), which
2
states that the family of processes n 0 (−1)N (n s) ds, where N is a standard Poisson process, converges in
law to a Brownian motion. The second one is constructed in terms of the kernels associated to the extension of
Donsker’s theorem to the plane.
This research work has been done in collaboration with Xavier Bardina and Maria Jolis (Universitat Autònoma
de Barcelona).

References
[1] D.W. Stroock. Lectures on topics in stochastic differential equations. With notes by Satyajit Karmakar. Tata
Institute of Fundamental Research Lectures on Mathematics and Physics, 68. Springer-Verlag, Berlin-New York,
1982.
[2] J.B. Walsh. A stochastic model of neural response. Adv. in Appl. Probab. 13 (1981), no. 2, 231-281.
Keywords : stochastic heat equation, weak convergence, two-parameter Poisson process, Donsker kernels
AMS Classification Numbers : 60B12, 60H15

12:40 Noèlia Viles, Universitat Autònoma de Barcelona

Continuity in the Hurst parameter of the law of the Russo-Vallois symmetric integral

Consider the laws in C([0, T ]) of the family of fractional Brownian motions

{B H , H ∈ (0, 1)} with Hurst parameter H ∈ (0, 1). It is easily seen that these laws converge weakly to
that of B H0 , when H tends to H0 ∈ (0, 1).

13
 It is interesting to study whether certain functionals of fractional Brownian motion conserve this property. That
is, we ask if their law (in C([0, T ])) remains near to that of the corresponding functional for B H0 when H is near
to H0 . This kind of result justifies the use of B Ĥ as a model in applied situations where the true value of the
Hurst parameter is unknown and Ĥ is some estimation of it.
We prove the convergence in law, in the space of continuous functions C([0, T ]), of the Russo-Vallois symmetric
integral of a non-adapted process with respect to the fractional Brownian motion with Hurst parameter H > 1/2
to the Russo-Vallois symmetric integral with respect to the fractional Brownian motion with parameter H0 , when
H tends to H0 ∈ [1/2, 1).

References
[1] M. Jolis and N. Viles, Continuity in the Hurst parameter of the law of the symmetric integral with respect to
the fractional Brownian motion, Preprint (2010).
[2] F. Russo and P. Vallois, Forward, backward and symmetric stochastic integration, Probab. Theory Related
Fields, no. 3, 97 (1993), 403–421.
Keywords : convergence in law, fractional Brownian motion, Russo-Vallois symmetric integral
AMS Classification Numbers : 60B12, 60J55, 60G15

13:15 – 14:30 Lunch

14:30 - 16:15 Invited Session: Stochastic Networks Venue: SSIU seminar hall
(Chair: Kavita Ramanan, Brown University)

14:30 Shankar Bhamidi, University of North Carolina, Chapel Hill

Flows, first passage percolation and random disorder in networks

Consider a connected network and suppose each edge in the network has a random positive edge weight.
Understanding the structure and weight of the shortest path between nodes in the network is one of the most
fundamental problems studied in modern probability theory and goes under the name first passage percolation.
It arises as a fundamental building block in many interacting particle system models such as the spread of
epidemics on networks. To a large extent such problems have been only studied in the context of the n-
dimensional lattice.
In the modern context these problems take on an additional significance with the minimal weight measuring the
cost of sending information while the number of edges on the optimal path (hopcount) representing the actual
time for messages to get between vertices in the network. Given general models of random graphs with random
edge costs, can one develop techniques to analyze asymptotics of functionals of interest which are robust to the
model formulation?
The aim of this talk is to describe a heuristic based on continuous time branching processes which gives very
easily, a wide array of asymptotic results for random network models in terms of the Malthusian rate of growth
and the stable age distribution of associated branching process. These techniques allow us to solve not only
first passage percolation problems rigorously but also understand functionals such as the degree distribution of
shortest path trees, congestion across edges as well as asymptotics for “betweeness centrality” a concept of
crucial interest in social networks, in terms of Cox processes and extreme value distributions. These techniques
also allow one to solve models of “weak disorder” in the context of the stochastic mean field model of distance,
a model of great interest in probabilistic combinatorial optimization.
Keywords : flows, weak disorder, first passage percolation, random graphs, stochastic mean field model of
distance, continuous time branching processes
AMS Classification Numbers : 60C05, 05C80, 90B15

15:05 Lea Popovic, Concordia University

Diffusion approximations for multiscale reaction networks

14
 Reaction networks in biological systems can involve population sizes with vastly differing abundances and
interactions whose rates also vary over several orders of magnitude. This wide variation in number and rate
yield phenomena that evolve on very different time-scales. These differing time-scales can be exploited to obtain
simplifications of complex models. In order to produce a reduced network one can use stochastic averaging
with respect to components whose relative change is faster and obtain a simplified system of equations for
the evolution of the components whose relative change is slower. In order to assess the fluctuations of the
reduced system, one has to take into account both intrinsic fluctuations as well as fluctuations around averaged
quantities. We present an appropriate diffusion approximation for the reduced reaction network.
Keywords : reaction networks, stochastic averaging, diffusion approximation, Poisson equation
AMS Classification Numbers : 60F05, 60F10, 60F17

15:40 Ayalvadi Ganesh, University of Bristol

Decentralised load balancing in closed and open networks

We study the performance of random load resampling strategies in parallel server systems. Clients initially
attach to an arbitrary server, but may switch server independently at random instants of time in an attempt to
improve their service rate. Load resampling is particularly relevant in scenarios where clients cannot predict
the load of a server before being actually attached to it. We derive tight estimates of the time it takes for a
given resampling strategy to achieve a perfect balance of the load across servers in a closed system. We also
study open systems where clients arrive according to a random process and leave upon service completion.
In this scenario, we characterize the stability region of various resampling strategies, and derive approximate
estimates of the sojourn time obtained by letting the number of servers grow large.
This is joint work with S. Lilienthal, D. Manjunath, A. Proutiere and F. Simatos.
Keywords : load balancing, randomised algorithms
AMS Classification Numbers : 60K25, 60K35, 68W15, 68W20

14:30 - 16:50 Contributed Session: Venue: Auditorium (main)

Interacting Particle Systems
(Chair: Maury Bramson, University of Minnesota)

14:30 Antar Bandyopadhyay, Indian Statistical Institute, Delhi

On the expected total number of infections for virus spread on a finite network

In this work we consider a simple virus spread model on a finite pop- ulation of nagents connected by some
neighborhood structure. Let G be the graph on n agents where an infection starts with some initial number of
infected sites. The infection spreads as follows: at each discrete time step, an infected vertex tries to infect its
neighbors with probability β ∈ (0, 1) independently of others and then it dies out. The process continues till all
infected sites dies out. We focus on obtaining proper lower bounds on the expected number of ever infected
sites. We obtain a simple lower bound when the infection starts with only one individual using breadth-first
search algorithm. We show that in a variety of examples this lower bound gives better approximation than some
of the known approximations through matrix-method based upper bounds. Moreover the lower bound works for
every value of β ∈ (0, 1). In fact, it is shown that if the graph G “locally looks like a tree” in the sense of the
local weak convergence then our lower bound is asymptotically exact. Finally, we also provide a generalization
of this bound when the virus spread starts with more than one infected agents.
This is joint work with Farkhondeh Sajadi.
Keywords : breadth-first search, local weak convergence, random graphs, spanning trees, virus spread
AMS Classification Numbers : 60K35, 05C80, 60J85, 90B15

15:05 Partha Sarathi Dey, University of California, Berkeley

Stein’s method for concentration inequalities - extension and applications

15
 Stein’s method is a semi-classical tool for establishing distributional convergence in problems involving complex
dependencies. A general way of deriving concentration inequalities using Stein’s method was introduced by
Sourav Chatterjee in his Ph.D. thesis. In this talk we present extension and some applications of Stein’s method
for concentration inequalities. We prove a concentration inequality for the magnetization in the Curie-Weiss
model at critical temperature where it obeys a non-standard normalization and super-Gaussian concentration.
We also show how this method can be used to derive exact large deviation asymptotics for the number of
triangles in the Erdós-Rényi random graph G(n, p) when p ≥ 0.31. This talk is based on joint work with Sourav
Chatterjee.
Keywords : Stein’s method, concentration inequality, Curie-Weiss model, random graph, large deviation
AMS Classification Numbers : 60E15, 60F10

15:40 Raazesh Sainudiin, University of Canterbury

Exact transition probabilities of the number of ancestral sample lineages in the discrete ancestral
recombination graph

We derive the exact one-step transition probabilities of the number of lineages that are ancestral to a random
sample from the current generation of a population that is evolving under the discrete Wright-Fisher model [1, 2]
for recombining haploid individuals [3, 4]. When we appropriately rescale time by the population size, our model
leads to the continuous time Markov chain called the ancestral recombination graph of Griffiths [4]. Explicit
discrete time computations of these probabilities by counting appropriate bipartite graphs for a time-reversed,
recombining, haploid, Wright-Fisher population are not available in the literature as most ARG models are
typically formulated approximately in the continuous setting of large population limits [3, 4]. Our discrete ARG
model provides a natural prior probability for statistical decision problems in conservation genetics of highly
threatened populations.
This is joint work with Bhalchandra Thatte.

References
[1] R. Fisher, The Genetical Theory of Natural Selection, Clarenson, Oxford, UK, 1930.
[2] S. Wright, Evolution in mendelian populations, Genetics 16 (1931), 97–159.
[3] R.R. Hudson, Properties of a neutral allele model with intragenic recombination, Theoretical Population
Biology 23 (1981), 183–201.
[4] R.C. Griffiths, The two-locus ancestral graph, Ishwar V. Basawa, and Robert L. Taylor, eds, Selected Pro-
ceedings of the Sheffield Symposium on Applied Probability: Held at the University of Sheffield, Sheffield,
August 16–19, 1989 (Hayward, CA: Institute of Mathematical Statistics, 1991), 100-117.
Keywords : discrete recombination, coalescent approximation, bipartite graphs, conservation genetics
AMS Classification Numbers : 92D15, 92D20, 92D40, 05A18, 60J80, 60J85

16:15 Rahul Roy, Indian Statistical Institute, Delhi

Learning from neighbours

Consider a discrete time interacting particle system on the integer line where particles are of two types– red and
blue. At each instance a particle tosses a coin and based on the success/failure of its own coin, as well as the
colours and tosses of the neighbouring coins it decides whether to retain its colour or change to the other colour.
We study the limiting behaviour of the resultant Markov chain. This model has been studied by economists to
model diffusion of technologies. This is joint work with Antar Bandyopadhyay and Anish Sarkar.
Keywords : particle systems, Markov chain
AMS Classification Numbers : 60K35

14:30 - 16:15 Contributed Session: Venue: Auditorium (seminar hall)

16
 Combinatorial Probability and Urn Models
(Chair: Srinivasa Varadhan, Courant Institute, New York University)

14:30 Ghurumuruhan Ganesan, Indian Statistical Institute, Delhi

Two properties of the random partition of an integer

In this paper, we consider two properties of integer partitions. It is well-known that there exists a one-to-one
correspondence between partitions of the integer n and the Ferrer diagram containing n points. In the first part
of our paper, we consider the number of ways a random Ferrer diagram containing n points can be dismantled.
In general d dimensions, this problem arises, for example, while obtaining the upper bound for the quantity
αd = lim sup nln(d−1)/d
pd (n)
where pd (n) is the number of d− dimensional integer partitions of n. Using the fact that
the number of planes needed to dismantle a d−dimensional Ferrer diagram is no more that dn1/d , an upper
√
bound on αd is obtained. For 2D and 3D integer partitions, we prove that almost always no more than (1.1) n
1
and (2.1)n 3 planes are respectively needed for dismantling a random Ferrer diagram. We think of the Ferrer
diagram as a collection of planes (lines) along X− and Y − axis and also obtain asymptotic expressions for the
minimum number of planes needed to dismantle in integer partition. We then extend the result to Plancherel
distributed and 3D integer partitions.
In the second part of our paper, we prove a conjecture of Yakubovich regarding limit shapes of “slices" of 2D
integer partitions of n when the number of summands m ∼ Anα . Doing so, we are also able to strengthen the
following result due to Erdös that concerns the relation between the number of integer partitions and composi-
tions. It is well-known that if pm (n) and p̃m (n) denote the number of partitions and compositions, respectively,
1
of an integer n into m summands, then, pm (n) ∼ p̃m (n) if m = o(n 3 ). As a corollary of our proof of the
1
Yakubovich conjecture, we prove that pm (n) ∼ p̃m (n) if and only if m = o(n 3 ). Finally, we extend Yakubovich’s
result on limit shapes of slices for the case of compositions for all values of multiplicities and all 0 < α < 1.
Keywords : dismantling Ferrer diagram, Yakubovich conjecture, random compositions
AMS Classification Numbers : 60C05

15:05 Krishanu Maulik, Indian Statistical Institute, Kolkata

Strong laws for balanced block triangular urns with irreducible diagonal blocks

We consider an urn model, whose replacement matrix is block upper triangular with irreducible diagonal blocks,
has all entries nonnegative and constant row sums. We obtain strong laws, under mild conditions, for the counts
of balls corresponding to each color. The analysis requires a rearrangement of the colors. We show that the
scalings for these counts depend on the Perron-Frobenius eigenvalues of a related diagonal block identified
through the rearrangement. We show that the limiting random variables corresponding to the counts of colors
within a block are constant multiples of each other. We provide an easy-to-understand explicit formula for them
as well, which involves the left eigenvector corresponding to the Perron-Frobenius eigenvalue of the diagonal
block used in obtaining the scale.
This is a joint work with Amites Dasgupta.
Keywords : urn model, balanced triangular replacement matrix, Perron-Frobenius eigenvalue, irreducible matrix
AMS Classification Numbers : 60G70, 60F05

15:40 Srinivasan Balaji, George Washington University

Phases in the diffusion of gases via the Ehrenfest urn

The Ehrenfest urn is a model for the mixing of gases in two chambers. We study the gradual change for an urn
containing n balls from the initial condition to the steady state. We look at the status of the urn after kn draws.
We identify three phases of kn : The growing sub linear, the linear, and the superlinear. In the growing sublinear
phase the amount of gas in either chamber is normally distributed, with parameters that are influenced by the
initial conditions. In the linear phase a different normal distribution applies, in which the influence of the initial
conditions is attenuated. The steady state is not a good approximation until a superlinear amount of time has
elapsed. At the superlinear stage the mix is nearly perfect, with a nearly perfect symmetrical normal distribution

17
 in which the effect of the initial conditions is completely washed away. We give interpretations for how the results
in different phases conjoin at the seam lines. The Gaussian results are obtained via martingale theory.
This work is jointly done with Hosam Mahmoud and Tong Zhang of George Washington University.
Keywords : urn model, random structure, martingale, central limit theorem
AMS Classification Numbers : 60C05, 60F05, 05A05, 60G42

16:15 – 17:00 Coffee Break

18
Monday, August 16, 2010

08:00 - 09:00 Registration

09:00 - 10:00 Plenary lecture: Erwin Bolthausen Venue: Auditorium (main)

(Chair: Rahul Roy, Indian Statistical Institute, Delhi)

Erwin Bolthausen, University of Zurich

Non-ballistic random walks in random environments

We report on a multiscale approach for the exit distributions of random walks in random environments, devel-
oped jointly with Ofer Zeitouni (University of Minneapolis, and Weizman Institute). The basic model is a random
walk moving in an environment generated by independent identically chosen transition probabilities, whose dis-
tributions are invariant under lattice isometries. The original approach required the dimension to be at least
three. We report also on more recent progress in dimension two, and about estimates for the Green’s function.
Keywords : random walks, random environments
AMS Classification Numbers : 60K37

10:00 - 11:00 Plenary lecture: Alison Etheridge Venue: Auditorium (main)

(Chair: Krishanu Maulik, Indian Statistical Institute, Kolkata)

Alison Etheridge, University of Oxford

Modelling evolution in a spatial continuum

One of the outstanding successes of mathematical population genetics is Kingman’s coalescent. This process
provides a simple and elegant description of the genealogical trees relating individuals in a sample of neutral
genes from a panmictic population, that is, one in which every individual is equally likely to mate with every
other and all individuals experience the same conditions. But real populations are not like this. Spurred on by
the recent flood of DNA sequence data, an enormous industry has developed that seeks to extend Kingman’s
coalescent to incorporate things like variable population size, natural selection and spatial and genetic structure.
But a satisfactory approach to populations evolving in a spatial continuum has proved elusive. In recent joint
work with Nick Barton, IST Austria, we introduced a framework for modelling the evolution of populations dis-
tributed in a spatial continuum. This leads to a new class of measure-valued processes which we will describe
and, as time permits, explore in this talk.
Keywords : spatial Lambda-Fleming-Viot processes, coalescent models, population modelling, spatial contin-
uum
AMS Classification Numbers : 60J25, 60J70, 92D10, 92D15

11:00 - 11:30 Coffee Break

11:30 - 13:15 Invited Session: Rough Path Analysis Venue: SSIU seminar hall
(Chair: Samy Tindel, University of Nancy)

11:30 Peter Friz, TU-Berlin and WIAS

A (rough) pathwise approach to SPDEs

We present a (rough)pathwise view on stochastic partial differential equations; examples range from the (fully
non-linear) stochastic HJB equation (introduced by Lions-Souganidis) to the (generalized) Zakai equation arising
from non-linear filtering (with correlation). Our results are based on the marriage of rough path analysis with
(2nd order) viscosity theory.
This is joint work with M. Caruana and H. Oberhauser.

19
 Keywords : rough paths, viscosity theory, SPDEs
AMS Classification Numbers : 35K55, 60H15

12:05 Yuzuru Inahama, Nagoya University

Laplace-type asymptotics in the rough path theory

In this talk we discuss Laplace-type asymptotic theorem for “small noise limit” of rough differential equations
driven by (1) fractional Brownian motion with Hurst parameter between 1/4 and 1/2 and (2) infinite dimensional
Brownian motion. This type of problems has a long history. It was initiated by Azencott in 1982 for SDEs in the
usual sense which is driven finite dimensional Brownian motion. The key of the proof is the Taylor expansion of
the Itô map in the rough sense around the point at which the minimum is achieved.
Keywords : rough path, Laplace approximation, large deviation, fractional Brownian motion
AMS Classification Numbers : 60F99, 60H10, 60G22

12:40 Samy Tindel, University of Nancy

Rough paths and Skorohod stochastic calculus for Gaussian processes

I will discuss some recent developments on rough path constructions for Gaussian processes with low regu-
larity, typically below the usual threshold of a Hölder exponent γ = 1/4. In spite of the fact that the iterated
integrals built up in this case cannot be considered as canonical, I will show that they give raise to reasonable
change of variable formulas, both in the Stratonovich and Skorohod sense. The method combines rough paths
considerations and Wick type calculus.
This talk is based on an ongoing work with Yaozhong Hu (Kansas) and Maria Jolis (Barcelona).
Keywords : rough paths analysis, Wick products, Malliavin calculus
AMS Classification Numbers : 60H07, 60H40, 60H05

11:30 - 13:15 Invited Session: Polymer Models Venue: Auditorium (main)

(Chair: Anish Sarkar, Indian Statistical Institute, Delhi)

11:30 Hubert Lacoin, Universita di Roma Tre

Directed polymers in random environment: Influence of disorder and the superdiffusivity phe-
nomenon

Directed polymers in random environment is a model to describe the behavior of polymer chain in a medium
with hetegeroneous impurities. The polymer configuration is modelized as directed random walk of N (N meant
to be a large integer) steps in Zd whose probability law is modified by random Gibbs weights. In high dimension,
this model exhibits a phase transition from a low-temperature phase where disorder localizes the polymer along
corridor of favorable environment to an high-temperature phase, where the polymer configuration keeps all the
characteristic features of the simple random walk (in particular: the invariance principle). The aim of this talk
is to give an overview of the model, explaining the crucial role played by the analysis of the partition function,
characterising the strong/weak disorder phase and giving an overview of recent superdiffusivity results.
Keywords : directed polymers, random media, superdiffusivity
AMS Classification Numbers : 82D60, 60K37, 82B44

12:05 Xia Chen, University of Tennessee

Brownian motion in a renormalized Poissonian potential

The model of Brownian motion in Poissonian potential describes a typical trajectory of a Brownian particle
surviving from being attracted by the obstacles randomly located in the space (think about the stars in the

20
 universe). In the existing literature, the random potential is defined as the convolution between a Poissonian
field and a bounded and locally supported function.
According to the NewtonâĂŹs law of universal attraction and some other related laws in physics, the most
natural way of constructing the random potential is to define it as the Riesz potential of the Poissonian field. On
the other hand, the Riesz potential of the Poissonian field blows up.
In this talk, this problem will be fixed by the way of renormalization. In addition, some asymptotic patterns of our
models will be established and more problems will be asked.
Part of the talk is based on some collaborative works with Kulic and Rosen.
Keywords : Brownian motion, Poissonian potential, Brownian potential
AMS Classification Numbers : 60K40, 60J45

12:40 Rongfeng Sun, National University of Singapore

Random walk among moving Poissonian traps

We review some old and new results on the survival probability of a random walk among a Poisson field of mov-
ing traps on Zd , which can also be interpreted as the solution of a parabolic Anderson model with a random
p time-
dependent potential. We show that the annealed survival probability decays asymptotically as exp(− 8t/π) for
d = 1, as exp(−πt/ log t) for d = 2, and as exp(−cd t) for some cd > 0 for d ≥ 3, while the quenched survival
probability always decays exponentially.
This talk is based on a survey article jointly with A. Drewitz, J. Gärtner, and A. F. Ramírez.
Keywords : Poissonian traps, parabolic Anderson model, directed polymer
AMS Classification Numbers : 60K35, 60K37

11:30 - 13:15 Contributed Session: Diffusions and Related Models Venue: Auditorium (seminar hall)
(Chair: Alison Etheridge, University of Oxford)

11:30 Vivek Shripad Borkar, Tata Institute of Fundamental Research

Small noise limits and Feller selection

A new selection principle is proposed for construction of a Feller solution to an ill-posed degenerate diffusion
process. This is based on constructing the Feller transition kernel from the unique (under suitable conditions)
continuous viscosity solutions to the backward Kolmogorov equation associated with the diffusion. The connec-
tions of this Feller process with the small-noise limits of nondegenerate perturbations of the diffusion and several
other related phenomena are also discussed. (Joint work with K. Suresh Kumar, Indian Institute of Technology,
Mumbai.)
Keywords : degenerate diffusions, Feller selection, small noise limit, viscosity solutions
AMS Classification Numbers : 60H10, 60J25, 34F05, 35K65

12:05 Krishnamurthi Ravishankar, SUNY-New Paltz

Marking the Brownain web and applications

In this talk I will discuss recent results (and some ongoing work) obtained in collaboration with C.M. Newman
and E. Schertzer. I will start with a brief introduction to the discrete web of coalescing simple random walks
and its continuum diffusive limit, the Brownian web (BW). After indicating how the continuum limit of the noisy
voter model (Glauber dynamics of nonzero temperature stochastic Ising model) is obtained using the marking of
(0, 2) (or bulk nucleation) points of the BW, the remainder of the talk will focus on (1, 2) points of the BW which
correspond to the boundary nucleation points of the stochastic Potts model. I will then de- scribe the marking
procedure for the (1, 2) points and indicate how it can be used to construct the Brownian net and then I will
describe how these marking(along with the markings of (0, 2) points) can be used obtain the continuum limit of
stochastic Potts model. If time permits I will discuss the marking construction of dynamical Brownian web.

21
 Keywords : coalescing random walks, Brownian web, Brownian net, Ising and Potts models, continuum limits
AMS Classification Numbers : 60K35, 60K35, 60F17

12:40 Siva Athreya, Indian Statistical Institute, Bangalore

Brownian motion on R trees

The real trees form a class of metric spaces that extends the class of trees with edge lengths by allowing
behavior such as infinite total edge length and vertices with infinite branching degree. We use Dirichlet form
methods to construct Brownian motion on a given locally compact R-tree equipped with a Radon measure ν .
We then characterize recurrence versus transience.
This is joint work with Anita Winter
Keywords : Brownian motion, R-trees
AMS Classification Numbers : 60J25, 60J75, 92B10

13:15 – 14:30 Lunch

14:30 - 16:15 Invited Session: Random Matrices Venue: Auditorium (main)

(Chair: Arup Bose, Indian Statistical Institute, Kolkata)

14:30 Steven J Miller, Williams College

Eigenvalue statistics for Toeplitz ensembles

Random Matrix Theory was developed by physicists to model the energy levels of heavy nuclei. We discuss
some of the classical results, and consider some ‘thin’ ensembles of real symmetric Toeplitz matrices. Such
matrices have far fewer degrees of freedom than the ensemble of all real symmetric matrices, and new behavior
is seen: the density of eigenvalues of normalized real symmetric matrices is a semi-circle; real symmetric
Toeplitz matrices is almost Gaussian. The deviations can be interpreted in terms of obstructions to Diophantine
equations. The obstructions vanish if we consider real symmetric palindromic Toeplitz matrices, in which case
the limiting spectral measure is a Gaussian. Using the method of moments and an analysis of the resulting
Diophantine equations and combinatorics, we show that the spectral measures converge almost surely. We end
with some results on what happens when we increase the palindromicity of the ensemble, in particular proving
this new ensemble has the fattest tails of any limiting measure observed to date.
Keywords : random matrix theory, Toeplitz ensembles
AMS Classification Numbers : 15B52, 60F05, 11D45

15:05 Manjunath Krishnapur, Indian Institute of Science

Limit of characteristic polynomials of a random matrix

We show that the characteristic polynomials of n × n Gaussian matrices, normalized by a random constant,
converges to a random analytic function. We show that the limit random analytic function is a mixture of Gaus-
sian analytic functions. The main feature of the proof is a version of Pólya’s urn scheme for Hilbert space valued
random variables.
This is joint work with Blint Virág.
Keywords : random matrix, random analytic function, gaussian analytic function
AMS Classification Numbers : 60B20, 30B20

15:40 Wenbo Li, University of Delaware

Probabilities of all real zeros for random polynomials

There is a long history on the study of zeros of a random polynomial whose coefficients are independent,

22
 identically distributed, non-degenerate random variables. We will provide exact and/or asymptotic bounds on
probabilities that a random polynomial under various distributions has all real zeros.
Keywords : random polynomials, probability estimates
AMS Classification Numbers : 60E05, 60G07, 60K40

14:30 - 16:15 Contributed Session: Venue: Auditorium (seminar hall)

Markov Models and Related Topics
(Chair: Louis Chen, National University of Singapore)

14:30 Tamal Banerjee, Indian Institute of Science

Pricing of defaultable bonds in a Markov modulated market

We address the problem of pricing defaultable bonds in a Markov modulated market. Using Merton’s structural
approach we show that various types of defaultable bonds are European type contingent claims. Thus pricing
a defaultable bond is tantamount to pricing a contingent claim in a Markov modulated market. Since the market
is incomplete we use the method of quadratic hedging and minimal martingale measure to derive locally risk
minimizing derivative prices, hedging strategies and the corresponding residual risks. By numerically solving
the appropriate PDEs with suitable boundary conditions we carry out an extensive numerical investigation of
defaultable bond prices and compare the credit spreads with some of the existing models.
Keywords : defaultable bond price, quadratic hedging, minimal martingale measures, credit spread
AMS Classification Numbers : 91B28, 91B70

15:05 Biswaroop Mookherjee, BASES The Nielsen Company

Markov chains with random transition probability matrix

A Markov Chain {St , t = 0, 1, . . .} with m discrete states in discrete time has Random Transition Probability
Matrix (TPM) T of order m × m. T can assume any of the matrices {T1 , T2 , . . . Tk } at a particular time. The
occurrence of Ti as TPM at time t depends only on what TPM was there at time t − 1. There is one more Markov
Chain {Mt , t = 0, 1 . . .} of these TPM’s i.e. the state-space of Mt is {T1 , T2 , . . . Tk }. The TPM of Mt is R of
order k × k .
In this paper we have studied limiting probabilities of the n-step TPM of St i.e. T n under different situations and
also we have found estimators of the parameters i.e. the elements of the matrices T1 , T2 , . . . , Tk and R. We
have first considered when Mt is observable and then also when Mt is hidden.
Let asij denotes the element at ith row and j th column of the TPM Ts . Evidently,
m
X
asij = 1,
j=1

for s = 1, . . . , k and i = 1, . . . , m. An important result we have found is that the n-step TPM T n converges as
n → ∞ if
a arij
Pm sij = Pm ,
i,j=1 asij i,j=1 arij
for all s, r, i, j.
Hence, the limiting probabilities depend on how transition probabilities of not being in the same state in the
matrix T change. We have shown application of the phenomena of this paper in agriculture and finance.
The first example taken is of a Markov Chain with two states denoting increase and decrease of salinity of soil
where the TPM is a random matrix and the probable matrices depend on a different Markov Chain i.e. of rainfall.
Here the states of both of the Markov Chains are observable. The other example taken is of a Markov Chain
with two states denoting increase and decrease of price of a stock where the TPM is a random matrix and the
probable matrices depend on a different Markov Chain i.e. of existence of bearish or bullish attitude among
investors. Here the price of the stock is observable but existence of bearish and bullish attitude is not.

23
 Keywords : Markov chains, random matrices, limiting probabilities, estimation
AMS Classification Numbers : 60J10, 60B20, 62M05

15:40 Subhamay Saha, Indian Institute of Science

Generalised semi-Markov processes

A semi-Markov process is completely determined by the holding time distributions in each state and the transition
probabilities of the embedded Markov chain. In this work we construct a generalized semi-Markov process
on a countable state space where the transition probabilities are age-dependent. Under certain condition we
establish the Feller property of the process. Finally we compute the limiting distribution of the process.
This is joint work with Mrinal Ghosh.
Keywords : transition rate function, Feller property, semi-Markov process, limiting distribution
AMS Classification Numbers : 60J27, 60J75

14:30 - 16:15 Contributed Session: Statistics and Simulations Venue: Auditorium (smu)
(Chair: Noam Berger, Hebrew University of Jerusalem)

14:30 Rahul Jain, University of Southern California

Solving Markov decision problems via simulations

It is well-known that solving Markov decision processes using dynamic programming is computationally in-
tractable. We propose a simulation-based framework that exploits the uniform laws of large numbers devel-
oped by Vapnik-Chervonenkis, and others. The Vapnik-Chervonenkis theory is a generalization of the classical
Glivenko-Cantelli theorem. They obtained necessary and sufficient conditions for a uniform law of large numbers
to hold for a class of measurable Boolean functions. This was later extended to bounded real-valued functions
by Pollard and others. It was shown that the rate of convergence depends on the epsilon-covering number of
the function class introduced by Kolmogorov and Tihomirov.
We present the beginnings of a corresponding empirical process theory for Markov decision processes. We pro-
vide uniform law of large number results for particular functionals of Markov decision processes. When uniform
convergence is obtained, we also obtain the rate of convergence in terms of P-dimension of the policy class.
Surprisingly, we find that how sample tra jectories of a Markov process are obtained from simulation matters for
uniform convergence: There are good simulation models (for which one may get uniform convergence) and bad
simulation models (for which one may not get uniform convergence for the same set of Markov processes). This
phenomenon seems to be the first such observation in the theory of empirical processes. Uniform convergence
results are also obtained for the average reward case, for some partially observed processes, and for Markov
games.
We then introduce a simulation-based framework for optimization of Markov decision processes. In particular,
we give an simulation-based algorithm to compute epsilon-optimal policy and show the regret minimization
property of such a framework. We then show its application to a class of Multi-armed bandit problems.
Keywords : Markov decision processes, simulation methods, empirical process theory
AMS Classification Numbers : 90, 90C40

15:05 Ananya Lahiri, Indian Institute of Technology Kanpur

Estimating the parameters of multiple chirp signals

Chirp signals occur quite naturally in different areas of signal processing. In this article, we consider the esti-
mation procedures of the different parameters of multiple chirp signals. Using the orthogonal structure of the
different components, we propose a sequential estimation procedure of the different parameters step by step.
It reduces to solve only a two dimensional optimization problem at a time. It is observed that the proposed
method produces the estimators which are asymptotically equivalent to the least squares estimators. We per-

24
 form some simulation experiments, to observe how the asymptotic results behave for small sample sizes, and
the performances are quite satisfactory. One data set has been analyzed for illustrative purposes.
Keywords : chirp signals, least squares estimators, strong consistency, asymptotic distribution, linear processes.
AMS Classification Numbers : 62M99

15:40 Sandeep Juneja, Tata Institute of Fundamental Research

A large deviations perspective on efficiency of simulation estimators

Consider a simulation estimator α(c) based on expending c units of computer time to estimate a quantity α. One
measure of (lack of) efficiency is the magnitude of the probability of making an error of (at least)  for a given (but
large) computer budget c, namely P (|α(c) − α| > ). We establish, in substantial generality, a large deviations
result that asserts that this probability decays roughly exponentially in the budget c. As a consequence, this point
of view leads to a preference for one estimator over another if its associated exponential decay rate constant
is the larger of the two. The decay rate depends critically on both the distribution of the estimator itself and
that of the random amount of computer time required to generate the estimator. Consequently, large deviations
analysis of simulation-based estimators that fail to account for the distribution of the random time required to
generate a sample may be misleading. This paper also develops an exact asymptotic for the error probability
in the important setting in which the estimator is constructed as an average of independent and identically
distributed samples. Our analysis also establishes that non-convex rate functions can arise in this context. The
principal results in this paper can be viewed as large deviations analogs to classical measures of simulation
efficiency going back to Hammersley and Handscomb.
This is a joint work with Peter Glynn, Department of Management Science and Engineering, Stanford University,
USA.
Keywords : large deviations, simulations, efficiency
AMS Classification Numbers : 60F10, 65C05

14:30 - 16:15 Contributed Session: Venue: SSIU seminar hall

Information Theory and Free Probability
(Chair: Alain-Sol Sznitman, ETH, Zurich)

14:30 Mokshay Madiman, Yale University

Concentration of information content for log-concave probability measures

Suppose a random vector X = X (n) taking values in Rn has a density f . The random variable

e
h(X) = − log f (X)

may be thought of as the (random) information content of X . The entropy of X is defined by h(X) =
R
− f (x) log f (x) dx = E e
h(X), when it exists.
Because of the relevance of the information content in various areas such as information theory, probability,
and statistics, it is intrinsically interesting to understand its behavior. In particular: Is it true that the information
content concentrates around the entropy in high dimension? This is clearly true when X = (X1 , . . . , Xn ) has
i.i.d. components; indeed, in that case, one also has that − n1 log f (X (n) ) → h(X1 ) almost surely, by the law of
large numbers, which is the simplest instance of the Shannon-McMillan-Breiman theorem for stationary, ergodic
processes.
In general, there is no reason for such a concentration property to hold. Our main purpose is to show, however,
that when the probability measure on Rn of interest is absolutely continuous and log-concave, log f (X) does
possess a powerful concentration property. Specifically, if X = (X1 , . . . , Xn ) is distributed according to a
log-concave density f on Rn , then
n √ o
Pr e
h(X) − h(X) ≥ t n ≤ 2 e−ct , (2)

25
 for all t > 0, where c ≥ 1/16 is a universal constant.
The result is also closely tied to the Shannon-McMillan-Breiman theorem [cf. Barron (1985) and Orey (1985) for
stationary, ergodic processes]. These works imply that if X = (X1 , X2 , . . .) is a stationary and ergodic process,
and X (n) = (X1 , . . . , Xn ), then, as n → ∞,

1
− log f (X (n) ) → h(X), a.s. (3)
n
where the “entropy rate” h(X) is defined by

1
h(X) := h(X (n) ). (4)
n
For non-stationary processes with arbitrary dependence, the entropy rate h(X) typically does not exist; so there
is no question of a statement like (2) holding. Nonetheless, (1) immediately yields the following extension of
the Shannon-McMillan-Breiman phenomenon (without a stationarity assumption): Suppose that X has a log-
concave distribution on R∞ with absolutely continuous finite dimensional projections. If the limit h(X) defined
by (3) exists, the property (2) holds. This is joint work with Sergey Bobkov at University of Minnesota.
Keywords : Keywords: concentration, entropy, log-concave, Shannon-McMillan-Breiman theorem
AMS Classification Numbers : AMS Classification Numbers: 60G07, 94A15

15:05 Rajat Subhra Hazra, Indian Statistical Institute, Kolkata

Subexponentiality of free regularly varying random variables

In this talk we consider {Xi }1≤i≤n to be free, identically distributed random variables affiliated to some W ∗
probability space with law having a regularly varying tail. We study the tail behaviour of the partial sums and
show that it is tail equivalent to its free maximum (in the sense of Ben Arous and Voiculescu (2006)). In particular,
we study the behavior of the remainder term in the expansion of Cauchy transform and Voiculescu transform
when the law has a regularly varying tail. The results also help us conclude that if additionally the law is infinitely
divisible then its free Lévy measure is regularly varying and the two laws are tail equivalent.
This is joint work with Krishanu Maulik.
Keywords : free probability, Cauchy transform, Voiculescu transform, regular variation
AMS Classification Numbers : 46L54, 60G70

16:15 – 17:00 Coffee Break

26
Tuesday, August 17, 2010

08:00 - 09:00 Registration

09:00 - 10:00 Plenary lecture: Louis Chen Venue: Auditorium Main

(Chair: Mrinal Kanti Ghosh, Indian Institute of Science)

Louis Chen, National University of Singapore

Normal approximation by Stein’s method

I will begin with Stein’s original ideas and discuss applications to local dependence, moderate deviations, dis-
cretized normal approximation and multivariate normal approximation. If time permits, I will touch on applications
to algebraic combinatorics and connection to the Malliavin calculus. This talk is based on material taken from a
book which I am writing jointly with Larry Goldstein and Qi-Man Shao.
Keywords : Stein’s method, normal approximation, moderate deviations, Malliavin calculus
AMS Classification Numbers : 60F05, 60F10, 60F17

10:00 - 11:00 Plenary lecture: S. R. S. Varadhan Venue: Auditorium Main

(Chair: Vivek Shripad Borkar, Tata Institute of Fundamental Research )

Srinivasa Varadhan, Courant Institute, New York University

Central limit theorems for dependent random variables

Martingale approximations have proved to be a very powerful method for establishing central limit theorems for
sums of dependent random variables under different mixing conditions. We will review them and describe some
modifications that enable to treat sums of the form
n
1 X
√ f (Xi , X2i , . . . , Xki )
n i=1

under suitable mixing conditions.

Keywords : central limit theorem, mixing
AMS Classification Numbers : 60F05, 60F17

11:00 - 11:30 Coffee Break

27
 Directory

Ankush Agarwal Indian Statistical Institute, Bangalore xchen@[Link]

Tata Institute of Fundamental Research speedwlk@[Link] Speaker p.-20
ankush@[Link]
Rajeev Bhaskaran Louis Chen
Siva Athreya Indian Statistical Institute, Bangalore National University of Singapore
Indian Statistical Institute, Bangalore brajeev@[Link] matchyl@[Link]
athreya@[Link] Chair p.-10 Speaker p.-27
Speaker p.-22
Abhay Gopal Bhatt Amites Dasgupta
Krishna B Athreya Indian Statistical Institute, Delhi Indian Statistical Institute, Kolkata
IMI and Indian Institute of Science abhay@[Link] amites@[Link]
kbathreya@[Link]
Anup Biswas Pierre Del Moral
Srinivasan Balaji TIFR Centre for Applicable Mathematics INRIA
George Washington University anup@[Link] [Link]-Moral@[Link]
balaji@[Link] Speaker p.-4 Speaker p.-5
Speaker p.-17
Arunangshu Biswas Freddy Delbaen
Antar Bandyopadhyay Presidency College ETH, Zurich
Indian Statistical Institute, Delhi arunb12002@[Link] delbaen@[Link]
antar@[Link] Speaker p.-9 Speaker p.-11
Speaker p.-15; Chair p.-1
Erwin Bolthausen Partha Sarathi Dey
Tamal Banerjee University of Zurich University of California, Berkeley
Indian Institute of Science eb@[Link] psdey1@[Link]
banerjee@[Link] Speaker p.-19; Chair p.-1 Speaker p.-15
Speaker p.-23
Vivek Shripad Borkar Santanu Dey
Saptak Banerjee Tata Institute of Fundamental Research TIFR, Mumbai
Indian Institute of Science borkar@[Link] dsantanu2002@[Link]
banerjeesaptak@[Link] Speaker p.-21; Chair p.-27
Yogeshwaran Dhandapani
Gopal K Basak Arup Bose Ecole Normale Superieure
Indian Statistical Institute, Kolkata Indian Statistical Institute, Kolkata yogesh@[Link]
gkb@[Link] bosearu@[Link] Speaker p.-7
Chair p.-10 Chair p.-22
Hugo Duminil-Copin
Gerard Ben Arous Maury Bramson University of Geneva
Courant Institute, New York University University of Minnesota [Link]@[Link]
gba1@[Link] bramson@[Link] Speaker p.-11
Speaker p.-2 Speaker p.-10
Richard Emilion
Noam Berger Wlodek Bryc University of Orleans
Hebrew University of Jerusalem University of Cincinnati [Link]@[Link]
berger@[Link] [Link]@[Link]
Speaker p.-1 Speaker p.-?? Alison Etheridge
University of Oxford
Shankar Bhamidi Arijit Chakrabarty etheridg@[Link]
University of North Carolina, Chapel Hill Indian Institute of Science Speaker p.-19
bhamidi@[Link] arijit@[Link]
Speaker p.-14 Xiao Fang
Xia Chen National University of Singapore
Suprio Bhar University of Tennessee g0700725@[Link]

28
 Speaker p.-7; Chair p.-10 Speaker p.-20
Peter Friz
TU-Berlin and WIAS Rahul Jain Ananya Lahiri
[Link]@[Link] University of Southern California Indian Institute of Technology Kanpur
Speaker p.-19 [Link]@[Link] ananya@[Link]
Speaker p.-24 Speaker p.-24
Ghurumuruhan Ganesan
Indian Statistical Institute, Delhi Mathew Joseph Wenbo Li
guru9r@[Link] University of Utah University of Delaware
Speaker p.-17 joseph@[Link] wli@[Link]
Speaker p.-22
Ayalvadi Ganesh Sandeep Juneja
University of Bristol Tata Institute of Fundamental Research Gabor Lugosi
[Link]@[Link] juneja@[Link] ICREA and Pompeu Fabra University
Speaker p.-15 Speaker p.-25 [Link]@[Link]
Speaker p.-10
Sunil Kumar Gauttam Naveen K P
Indian Institute of Technology Bombay Indian Institute of Science Jan Maas
sunil@[Link] naveenkp@[Link] Universität Bonn
maas@[Link]
Subhroshekhar Ghosh Sameer Kamal Speaker p.-6
University of California, Berkeley Tata Institute of Fundamental Research
subhroshekhar@[Link] [Link]@[Link] Mokshay Madiman
Yale University
Mrinal Kanti Ghosh Antti Kemppainen [Link]@[Link]
Indian Institute of Science University of Helsinki Speaker p.-25; Chair p.-5
mkg@[Link] [Link]@[Link]
Chair p.-27 Speaker p.-12 D. Manjunath
Indian Institute of Technology Bombay
Suman Guha Mustafa Khandwawala dmanju@[Link]
Indian Statistical Institute, Kolkata Indian Institute of Science
bst0404@[Link] mustafa@[Link] Deemat C Mathew
Indian Statistical Institute
Martin Hairer Arturo Kohatsu Higa deemat8r@[Link]
University of Warwick Osaka University
[Link]@[Link] arturokohatsu@[Link] Jonathan Mattingly
Chair p.-6 Speaker p.-3; Chair p.-2 Duke University
jonm@[Link]
Rajat Subhra Hazra Manjunath Krishnapur Speaker p.-3
Indian Statistical Institute, Kolkata Indian Institute of Science
rajat_r@[Link] manju@[Link] Krishanu Maulik
Speaker p.-26 Speaker p.-22; Chair p.-11 Indian Statistical Institute, Kolkata
krishanu@[Link]
Kenneth Hochberg K. Suresh Kumar Speaker p.-17; Chair p.-19
Bar-Ilan University Indian Institute of Technology Bombay
hochberg@[Link] suresh@[Link] ely Merzbach
Speaker p.-11 Bar-Ilan University
Yuzuru Inahama merzbach@[Link]
Nagoya University M. Ashok Kumar
inahama@[Link] Indian Institute of Science Steven J Miller
Speaker p.-20 ashokm@[Link] Williams College
sjm1@[Link]
Srikanth K Iyer Hubert Lacoin Speaker p.-22
Indian Institute of Science Universita di Roma Tre
skiyer@[Link] lacoin@[Link] Amit Kumar Misra

29
Indian Institute of Technology Kanpur Lluís Quer-Sardanyons
amishra@[Link] Universitat Autònoma de Barcelona Koushik Saha
quer@[Link] Bidhannagar College
Andrea Montanari Speaker p.-13 koushiksaha877@[Link]
Stanford University Speaker p.-8
montanari@[Link] Kavita Ramanan
Speaker p.-1, 5 Brown University Raazesh Sainudiin
Kavita_Ramanan@[Link] University of Canterbury
Biswaroop Mookherjee Speaker p.-5; Chair p.-14 rs228@[Link]
BASES The Nielsen Company Speaker p.-16
biswaroopm@[Link] Sundareswaran Ramasubramanian
Speaker p.-23 Indian Statistical Institute, Bangalore Marta Sanz-Solé
ram@[Link] Facultat de Matemàtiques
Leonid Mytnik [Link]@[Link]
Technion - Israel Institute of Technology Manikandan Rangaswamy
leonid@[Link] CUSAT Anish Sarkar
Speaker p.-6 mani552@[Link] Indian Statistical Institute, Delhi
[Link]@[Link]
Shobhit Nigam Bhamidi Visweswara Rao Speaker p.-6; Chair p.-20
Indian School of Mines Chennai Mathematical Institute
shobhitngm@[Link] bhamidivrao@[Link] Evelina Shamarova
Chair p.-1 University of Porto
Pierre Nolin evelinas@[Link]
Courant Institute, New York University K.S. Mallikarjuna Rao Speaker p.-12
nolin@[Link] Indian Institute of Technology Bombay
Speaker p.-12 [Link]@[Link] Vladas Sidoravicius
Speaker p.-4 CWI/IMPA
Daniel Paulin [Link]@[Link]
National University of Singapore Sreenivasan Ravi Speaker p.-2
paulindani@[Link] University of Mysore
sreenivasanravi@[Link] Chandramani Singh
Giovanni Peccati Speaker p.-8 Indian Institute of Science
Luxembourg University chandra@[Link]
[Link]@[Link] Krishnamurthi Ravishankar
Speaker p.-2 SUNY-New Paltz Jeetendra Bahadur Singh
ravi@[Link] Indian School of Mines
Alberto Algergo Pinto Speaker p.-21 jeetendra01@[Link]
University of Porto
aapinto1@[Link] Sharada Nandan Raw Vinit Kumar Sinha
Speaker p.-3 Indian School of Mines Madurai Kamraj University
shardaraw@[Link] svinit83@[Link]
Lea Popovic
Concordia University Parthanil Roy Karthik Sriram
lpopovic@[Link] Michigan State University Indian Institute of Management
Speaker p.-14 roy@[Link] karthik.sriram09@[Link]

Shakti Prasad Rahul Roy Rongfeng Sun

Indian School of Mines Indian Statistical Institute, Delhi National University of Singapore
[Link]@[Link] rahul@[Link] matsr@[Link]
Speaker p.-16; Chair p.-19 Speaker p.-21
Jeremy Quastel
University of Toronto Subhamay Saha Rajesh Sundaresan
quastel@[Link] Indian Institute of Science Indian Institute of Science
Speaker p.-6 subhamay585@[Link] rajeshs@[Link]
Speaker p.-24

30
Alain-Sol Sznitman Speaker p.-20; Chair p.-19 Indian Institute of Science
ETH, Zurich vsrini@[Link]
sznitman@[Link] Srinivasa Varadhan
Speaker p.-1 Courant Institute, New York University Noèlia Viles
varadhan@[Link] Universitat Autònoma de Barcelona
Debleena Thacker Speaker p.-27 nviles@[Link]
Indian Statistical Institute, Delhi Speaker p.-13
thackerdebleena@[Link] Rasbagh Vasudeva
Speaker p.-7 University of Mysore Hoi Ying Wong
[Link]@[Link] The Chinese University of Hong Kong
Samy Tindel Speaker p.-8 hywong@[Link]
University of Nancy Speaker p.-10
tindel@[Link] Srinivasan Venkatramanan

31
 Conference Venue
Talks:- Plenary lectures will be held in the main auditorium. Invited and Contributed Sessions will
be held in one of four venues (as indicated in the program): Auditorium (main), Auditorium(seminar),
SSIU Seminar hall and Auditorium (SMU). The Auditorium(main) and Auditorium(seminar) are located
in the second floor of the Main auditorium and the SSIU Seminar hall and Auditorium (SMU) are located
in the Administration. (Please see map below)

Schematic Map of the Indian Statistical Institute, Bangalore Centre (Not to Scale) Schematic Map of Second Floor of Main Building
(Not to Scale)

Offices Offices

Residence
Stairs
Offices Statmath Offices
Office
Residence
Path through the woods

Residence

Rest Room Offices

Main Stairs
PCO
Hostel

Auditorium
Offices

Auditorium

(SMU)
Offices Offices
Canteen

Main Building

Hostel SSIU
Guest House
Rest Room Offices (Seminar Hall)

Registration Desk: It will be in the foyer of the main auditorium building. This desk will also serve
as an information desk for general queries and distribution of announcements.

Onsite Registration: Participants who have arranged to pay for congress registration by cash
should do so at the Registration Desk.

Meals: Breakfast on all days and traditional south Indian dinner event (15th evening) will be served in
the institute canteen. The food (primarily vegetarian) for Tea/Coffee breaks, Lunches, welcome reception
(13th evening), will be served in the main auditorium building. If you have any dietary restrictions then
please contact the registration desk.

1
 Emergency Contact numbers: Siva Athreya – 98862 93244, Statmath unit office – (within campus)
440, 28482724, and Security officer – (within campus) 502, 9741874390.

Medical Emergency: Incase of a medical emergency please contact one of the numbers mentioned
above. For general consultation, a doctor visits the campus between 2-4pm on weekdays. The nearest
hospital is Sahana Hospital, Kengeri Satellite Town, Kengeri. Ph: 28482806 / 28485806

Getting to Airport or Train station from Conference Venue: The easiest way to the airport
is to hire a taxi. There are several taxi-service providers in Bangalore. The two main ones that provide
service on meter to the airport are MERU Cabs (44224422) or Easy Cabs (43434343).
The easiest way to get to the train station is to take 222A bus from the stop on Mysore road. You
can also hire a taxi from Aries cabs (26606744) or Spot Taxi Service (4110 0000) or City Safari (23225522,
23238899).

Computing Facilities:
Our computer centre has Linux PCs and WINDOWS-Vista enabled PCs. The computer centre will remain
open on 13th, 16th and 17th from 09:00 Hrs. to 17:30 Hrs. It is located on the first floor of the main
building.
WiFi: The institute guest house, auditorium and the main building are WiFi enabled. You can connect
your laptop to the network with the network key ISIbc.

Daily Announcements and Program Changes: An information sheet containing last minute
announcements, program changes and other information would be available at the Registration desk and
be posted on the notice board.

Library: The institute’s library is located in the ground floor of the main building. The opening hours
are from 9:30am to 5:30pm.

Phone: A couple of Public Call Operators (PCO) phone stands are in front of our institute’s main
gate. At such PCO local, interstate and international calls can be made on a payment basis. (See map
for location)

Bank Services and Foreign Exchange: There are several ATM’s in the vicinity of the Institute.
Sadly, not within walking distance. We have arranged for a Money Exchange agency to be present at the
conference venue during the first and the fourth day of the meeting from 11am to 3pm.

2
 Social Events

Welcome Reception : 13th August, 2010 There will be welcome reception from 17:00pm to 18:00pm at
the main auditorium. Soft drinks, Tea, Coffee, variety of finger foods, cakes and north Indian chat items
will be served.

Special Reception: 14th August 2010 This will be held at The club (situated near the institute) from
17:00pm to 18:30pm. Soft drinks, Alcoholic beverages (cash bar), Tea, Coffee, variety of finger foods, cakes
and north Indian chat items will be served. There will be a cash bar.

You will need to sign up for the following events.

You can do so online or on sign up sheets at the registration desk.

Traditional South Indian Dinner: 15th August 2010 This will be held a the main canteen of the institute
from 20:00pm. This traditional south-Indian meal will be served to you on a banana leaf and will be a
vegetarian affair.

The Green Heritage Walk in Lalbagh : 15th August 2010 A Nature Walk in Lalbagh Gardens, The walks
start at 7am and end with breakfast at 1030am. (See [Link] for more details)

Bangalore through the Ages Bus Tour : 15th August 2010 a 4hourcity tour that uses a bus and has
several short stops. It is not a sightseeing tour, but a themed tour that covers elements of all Banglaore
walks traditional, colonial, green and contemporary. It will start at 9:30am and end with lunch.(See
[Link] for more details)

Trip to Belur-Halibed: 15th August 2010 Belur and Halibed are two small towns in Hassan district
of Karnataka. The tour will take you to visit the Hoysala temples built in the 12th century (See
[Link] for more details). The trip will start at 6:30am and
you will return by 6pm.

Conference Dinner: 16th August 2010 This will be held at the restaurant Sunny’s from 19:00pm to
21:00pm. The restaurant is located on Vital Mallya road, near UB city next to the Shell petrol station.
You will need to sign up for this event. Sign up sheets are at the registration desk

3
 Lectures in Probability and Stochastic Process Series V

Venue: Indian Statistical Institute, Bangalore

Dates: December 23,24,26,27, 2010

Speakers:

Krishanu Maulik, Indian Statistical Institute, Kolkata

Heavy Tailed Distributions

S. R. S. Varadhan, Courant Institute, New York University

Schramm Loewner Evolution

Each speaker will give eight 50 minute lectures. These will be supplemented with eight discussion/open
problem sessions. Please visit [Link] for more details.

Organisers: Siva Athreya (Indian Statistical Institute, Bangalore), Antar Bandhyopadyay (Indian Sta-
tistical Institute, Delhi), Krishanu Maulik (Indian Statistical Institute, Kolkata). If you wish to attend
then please send an email to one of the organisers.

History and Scope: Lectures on Probability and Stochastic Processes Series, is an annual event and,
was started in 2006. It is designed to have two mini-courses on topics of recent interest given by experts
in the respective areas. These courses are expected to expose the participants to the important problems
being considered in the current literature, as well as the available techniques to work with them. It is
expected that this gathering will initiate collaborative work across different research centres of India and
elsewhere. To facilitate this activity, there is sufficient time for sessions, where researchers can pose and
discuss problems, which can be of interest to others in the audience.