0% found this document useful (0 votes)
7 views49 pages

Hill Climbing for 4-Queens Problem

The document provides an introduction to Artificial Intelligence (AI) and Machine Learning (ML), focusing on local search algorithms and optimization problems. It discusses the transformation of problems from satisfaction to optimization, various search techniques like hill climbing, simulated annealing, and genetic algorithms, and their applications in solving problems such as the N-Queens problem. Additionally, it covers constraint satisfaction problems and knowledge representation in AI, highlighting the importance of effective search strategies and representation in reducing search space.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
7 views49 pages

Hill Climbing for 4-Queens Problem

The document provides an introduction to Artificial Intelligence (AI) and Machine Learning (ML), focusing on local search algorithms and optimization problems. It discusses the transformation of problems from satisfaction to optimization, various search techniques like hill climbing, simulated annealing, and genetic algorithms, and their applications in solving problems such as the N-Queens problem. Additionally, it covers constraint satisfaction problems and knowledge representation in AI, highlighting the importance of effective search strategies and representation in reducing search space.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

1

Introduction to
AI and ML

 Instructors:
 Dr. Nishant Gupta & Dr. Manoj Kumar
Local Search 2

 Difference in problem Formulation


 Past Discussions: path to a goal is solution to
problem
 A State/node was some partial information
 systematic exploration of search space (DFS,
BFS, etc.)
 Now, not the path but a state is solution
to problem
 Search is in solution space
 For some problems path is irrelevant. –
E.g., 8-queens
 Different algorithms can be used
 Local Search
Local Search 3

 Satisfaction – Find me a path to Goal


 Goal Satisfaction (Optimal or not)
 Reach the goal node
 Constraint satisfaction
 What is optimization problem?
 Optimization
 optimize(objective function) Constraint Optimization
 Transformation of problems from Satisfaction to the Optimization
problem and Back Forth - Modelling
Local Search and Optimization 4

 Local search
 Keep track of single current state
 Move only to neighboring states
 Ignore paths
 Advantages: –
 Use very little memory
 Can often find reasonable solutions in large or infinite (continuous) state
spaces
 “Pure optimization” problems –
 All states have an objective function
 Goal is to find state with max (or min) objective value
 Does not quite fit into path-cost/goal-state formulation
 Local search can do quite well on these problems.
Example: N Queens 5

 Put n queens on an n x n board with no two queens on the same


row, column, or diagonal

 Is this a satisfaction or optimization problem?


Hill Climbing Search: 8 Queens 6

 Need to convert to an optimization problem


Optimization: N Queens 7

 Appropriate objective function?


 Maximum number of pairs that can attack each
other?
 Minimum number of pairs that can attack each
other?
 Do we need maximization or minimization
problem?
 h (Objective function) = number of pairs of
queens that are attacking each other
 h = 17 for the above state
Search Space 8

 State
 All 8 queens on the board in some configuration
 How many states or solutions?
 Local search – we search among the solution space
 Successor function/ Neighborhood function – We have to define
 move a single queen to another square in the same column
 Many others
 Example of a objective function h(n): –
 the number of pairs of queens that are attacking each other
 (so we want to minimize this)

Now we have converted the N Queens problems into a local search formulation
Hill climbing search: 8 Queens 9

 Is this a solution?
 What is h?
Trivial Algorithms 10

 Random Sampling
 Generate a state randomly
 Random Walk
 Randomly pick a neighbor of the current state
 Both algorithms asymptotically complete.
Hill-climbing (Greedy Local Search) 11
max version
Hill-climbing search 12

 “a loop that continuously moves towards increasing value”


 terminates when a peak is reached
 Aka greedy local search
 Value can be either
 Objective function value
 Heuristic function value (minimized)
 Hill climbing does not look ahead of the immediate neighbors
 Can randomly choose among the set of best successors
 if multiple have the best value
Landscape of search 13

 Hill Climbing gets stuck in local minima depending on?


Hill-climbing on 8-queens 14

 Randomly generated 8-queens starting states


 14% the time it solves the problem
 86% of the time it get stuck at a local minimum
 However
 Takes only 4 steps on average when it succeeds
 And 3 on average when it gets stuck
 (for a state space with 8^8 =~17 million states)
Hill Climbing Drawbacks 15

 Local Minima

 Plateau

 Diagonal Ridges
Escaping Shoulders: Sideways Move 16

 If no downhill (uphill) moves, allow sideways moves in


hope that algorithm can escape
 Need to place a limit on the possible number of sideways
moves to avoid infinite loops
 For 8-queens
 Now allow sideways moves with a limit of 100
 Raises percentage of problem instances solved from 14 to
94%
 However….
 21 steps for every successful solution
 64 for each failure

Problem?
Tabu Search 17

 prevent returning quickly to the same state


 Keep fixed length queue (“tabu list”)
 add most recent state to queue; drop oldest
 Never make the step that is currently tabu’ed
 Properties: –
 As the size of the tabu list grows, hill-climbing will asymptotically
become “non-redundant” (won’t look at the same state twice)
 In practice, a reasonable sized tabu list (say 100 or so) improves
the performance of hill climbing in many problems
Escaping Shoulders/local Optima 18
Enforced Hill Climbing
 Perform breadth first search from a local optima
 to find the next state with better h function
 Typically,
 prolonged periods of exhaustive search
 bridged by relatively quick periods of hill-climbing
 Middle ground b/w local and systematic search
Hill Climbing: Stochastic Variations 19

 Stochastic hill-climbing :
 Random selection among the uphill moves.
 The selection probability can vary with the steepness of the uphill
move.
 To avoid getting stuck in local minima
 Random-walk hill-climbing
 Random-restart hill-climbing
 Hill-climbing with both
Hill Climbing with random walk 20

 When the state-space landscape has local minima, any search that
moves only in the greedy direction cannot be complete
 Random walk, on the other hand, is asymptotically complete

Idea: Put random walk into greedy hill-climbing


 At each step do one of the two
 Greedy: With prob p move to the neighbor with largest value
 Random: With prob 1-p move to a random neighbor
Hill Climbing with random restarts 21

 If at first you don’t succeed, try, try again!


 Different variations
 For each restart: run until termination vs. run for a fixed time
 Run a fixed number of restarts or run indefinitely
 If you want to pick one local search algorithm, learn this
one!!
Hill Climbing with Both 22

 At each step do one of the three


 Greedy: move to the neighbor with largest value
 Random Walk: move to a random neighbor
 Random Restart: Resample a new current state
Simulated Annealing 23

 Simulated Annealing = physics inspired twist on random walk


 Basic ideas:
 like hill-climbing identify the quality of the local improvements
 instead of picking the best move, pick one randomly
 say the change in objective function is delta (d)
 if d is positive, then move to that state
 otherwise:
 move to this state with probability proportional to d
 thus: worse moves (very large negative d) are executed less often
 over time, make it less likely to accept locally bad moves
Simulated Annealing 24
Temperature T 25

 High T: probability of “locally bad” move is higher


 Low T: probability of “locally bad” move is lower
 Typically, T is decreased as the algorithm runs longer
 i.e., there is a “temperature schedule”
Local Beam Search 26

 Idea: Keeping only one node in memory is an extreme reaction to memory


problems.
 Keep track of k states instead of one
 Initially: k randomly selected states
 Next: determine all successors of k states
 If any of successors is goal finished
 Else select k best from successors and repeat

 Problem: quite often, all k states end up on same local hill


Genetic Algorithms 27

 Twist on Local Search: successor is generated by combining two parent states


 A state is represented as a string over a finite alphabet (e.g. binary) – String
representation
 8-queens - State = position of 8 queens each in a column – String represent the State
 Start with k randomly generated states (population)
 Evaluation function (fitness function):
 Higher values for better states.
 e.g., # non-attacking pairs in 8-queens
 Produce the next generation of states by “simulated evolution”
 Random selection
 Crossover
 Random mutation
Genetic Algorithms 28

Has the effect of “jumping” to a completely different new part of the search
space (quite non-local)
Genetic Algorithms 29

 Positive Points
 Random exploration can find solution that local search cannot (via
crossover)
 Negative Points
 Large number of tunable parameters
 Useful on some set of problems but there is no proof that GA is better
than hill climbing with random restarts
Definition of AI 30

 Up until now

“AI is Search”
Definition of AI 31

 Up until now

“AI is Search”

“AI is Representation”
Definition of AI 32

 Up until now

“AI is Search”

“AI is Representation”
Representation reduces the total search space
Constraint Satisfaction Problems (CSPs) 33

 Standard search problem:


 state is a “black box”—any old data structure
 that supports goal test, eval, successor
 CSP:
 state is defined by variables X_i with values from domain D_i
 goal test is a set of constraints specifying allowable combinations of values for
subsets of variables
 Simple example of a formal representation language
 Allows useful general-purpose algorithms with more power than standard
search algorithms
Constraint Satisfaction Problems (CSPs) 34
Constraint Satisfaction Problems (CSPs) 35
Constraint Graph 36

Binary CSP: each constraint relates at most two variables

Constraint graph: nodes are variables, arcs show constraints


Variety of Constraints 37
Real world CSPs 38

• Timetabling problems e.g., which class is offered when and


where?
• Hardware configuration
• Spreadsheets
• Transportation scheduling
• Factory scheduling
• Floor planning

Notice that many real-world problems involve real-valued variables


Backtracking Search 39
Backtracking Search 40
Backtracking Search 41
Improving Backtracking 42
Improving Backtracking 43
Improving Backtracking 44

This is variable order heuristic


Improving Backtracking 45
Forward Checking 46
Knowledge Representation 47

• represent knowledge about the world in a manner that facilitates


inferencing (i.e. drawing conclusions) from knowledge.

• Example: Arithmetic logic – x >= 5

• In AI: typically based on


• Logic – medium of conversation with the AI system
• Probability
• Logic and Probability
Knowledge Representation 48

• Basic idea of Logic – From true assumptions deduce true conclusions


KR Languages 49

• Propositional Logic
• Predicate Calculus
• Frame Systems
• Rules with Certainty Factors
• Bayesian Belief Networks
• Influence Diagrams
• Ontologies
• Semantic Networks
• Concept Description Languages
• Non-monotonic Logic

You might also like