Section overview graph for '(Shallow) Unsupervised Learning' by Prof.

Laurenz Wiskott
(solid and dashed arrows indicate required and recommended sections, respectively)
(background: blue = mathematical theory; green = algorithms; yellow = applications; white = external)
(red text are hyperlinks and lead to further information; numbers in parentheses indicate estimated time in minutes)

Machine Learning Basics (2 lectures)

1 Learning paradigms (45) 2 Training and testing (21)

3 Shallow unsupervised Almost all units
methods (96) in this graph

Independent Component Analysis (2 lectures)

Mathematics:
4 FastICA (40)
Functions 5

5 Application (6)

Mathematics: 3 Cumulant based

1 Intuition (58) 2 Cumulants (54)
Bayesian Theory 1 ICA (CuBICA) (32)

Reinforcement Learning

Principal Component Analysis (2 lectures) Slow Feature Analysis (1 lecture, including nonlinear expansion)
Mathematics: 2 Applications of SFA
1 Intuition (60) 2 Formalism (124) 1 SFA algorithm (41) in machine learning (30)
Linear Algebra 7

3 Applications + (12) Nonlinear Expansion (0.2 lecture)

1 Nonlinear expansion (19)

Laplacian Matrix (2 lectures)

5 Locality Preserving
Projections (LPP) (13)

Mathematics: 1 Heat diffusion 4 Laplacian

2 Formalism (68) 3 Similarity graphs (8)
Differential Equations 1 analogy (43) Eigenmaps (LEM) (25)

6 Spectral Clustering (15)

Self-Organizing Maps (0.5 lecture)

1 Algorithm (25) 2 Applications (13)

Clustering (1 lecture)
Mathematics: Vector Quantization (1 lecture) 2 Hard partitional
Functions 5 clustering
1 Batch learning 2 Online learning (k-Means) (20)
algorithms (26) algorithms (35)
Mathematics:
Linear Algebra 2 5 Agglomerative
1 Overview (7) hierarchical clustering
(Dendrogram) (28)

4 Density based
clustering
(DBSCAN) (8)

3 Soft partitional
clustering
(Gaussian mix. mod.) (38)

Graphical Models (3 lectures)

Mathematics: 5 Bayesian learning for
Bayesian Theory 1 binary variables (50)

2 Application: Image 6 Learning in

super-resolution (18) Bayesian networks (15)

1 Inference in 3 Inference in
Gibbsian networks + (68)
Bayesian networks (72) Restricted Bolzmann Machines (1 lecture)
Mathematics: 4 Approximate 1 Binary Restricted
2 Applications (7)
Stochastics inference (21) Boltzmann Machines (69)