i CSE 471/598 Lecture Notes (Spring 2009)

CSE 471/598 Lecture Notes (Spring 2009)

Motto: A day spent staring at the powerpoint slides and listening to the .WAV files by yourself
can save as much as 75 min in the class with Rao

Week 1: Intro; Intelligent agent design [R&N Ch 1, Ch 2]

Week 2: Problem Solving Agents [R&N Ch 3 3.1--3.5]

Audio of [Jan 27, 2009]. goal-based agent used to give an overview of why various topics will be covered in the class; problem solving agent, and different types of problems including single-state; multi-state; contingent and exploration--and their relation to deterministic, inaccessible, non-deterministic and open (map-less) worlds respectively; start of single-state problem solving; 8/15/24 puzzles.
Audio of [Jan 29, 2009]. Discussion of the "medicate-without-killing" problem--and sensing and non-deterministic actions in beleif space. Formulating search algorithms; properties of search algorithms; discussion of bfs, dfs, dlfs and iddfs and their properties. Graphsearch and handling duplicate nodes.

Week 3: Informed search [R&N Ch 3 3.1--3.5]

Week 4: CSPs and Local Search[R&N Ch 5.1--5.3; Ch 4 4.3]

Week 5: Local Search and Propositional Logic[R&N Ch 4 4.3; Ch 7.1--7.6]

Week 6: Propositional Logic --> Plausible reasoning[R&N Ch 7.1--7.6; [ch 13 13.1--13.5]]

Week 7: Representations for Reasoning with Uncertainty[ch 13 13.1--13.5]]

Week 8: Bayes Nets: Specification & Inference[ch 13 13.1--13.5]]

[Mar 19, 2009] Bayes Nets--Specification, Independence Assertions, Ways of constructing bayes nets (when domain causality is known), approaches for increasing network sparsity and reducing CPT size (1) Audio of first few min (2) Audio of the remaining 1hour 5min or so of the class

Week 9: Bayes Nets: Inference[ch 13 13.1--13.5]]

(Here is a fully worked out example of variable elimination)

Week 10: Sampling methods for Bayes net Inference; First-order logic start[ch 13.5; ]

Week 11: Unification, Generalized Modus-Ponens, skolemization and resolution refutation.

Week 12: Reasoning with change-->Planning

Week 13: Planning, MDPs

Week 14: Learning

Audio of [Apr 30, 2009] Review of min-max search and alpha-beta pruning; handling computational limitations by doing limited-depth min-max search. Notion of evaluation functions (and their connection to heuristics). Discussion of deep-blue. Discussion of game-tree pathology (and understanding why going deeper actually does help, considering that the accuracy of the evaluation function is not necessarily a function of depth). Idea of quiscence search. Idea of horizon effect (and the stupidity of mentioning the pre-specified depth of your search to your opponent; McCain's arguments against making public declarations about end to Iraq presence).
How to learn evaluation functions? Seague into learning. Idea of classification learning. Idea of representation for the hypotheses. The bruteforce view of classification learning as picking the hypothesis that best matches the training data. The size of the hypothesis space.

Week 15: Learning (decision trees; learning curves etc); swan song

Audio of [May 5, 2009]> Learning vs. Compression; Bias/Variance tradeoff (and fitting training data vs. test data). Decision tree learning as picking the tree that fits that data the best. Notion of false positives and false negatives--and cost-based learning. Greedy learning of decision trees. Selecting the feature to split on. Entropy as a measure. Performance of decision trees in Rusell domain vs. Majority rule.
A brief discussion of naive bayes classifier and comparison of its performance to decision tree learner.

Subbarao Kambhampati

Last modified: Sat May 9 12:20:22 MST 2009