Review of AI Professor: Liqing Zhang Contact Information: Email: zhang-lq@cs.sjtu.edu.cn Tel: 6293 2406
Chapter 1 Basic concepts: What is AI? Comparison between Human Intelligence and AI Milestones of AI
Chapter 2: Stimulus-Response Agents 2.1 Perception and Action Perception Action Boolean Algebra Clauses and Forms of Boolean Functions 2.2 Representing and Implementing Action Functions Production Systems Networks The Subsumption Architecture
Chapter 2: Stimulus-Response Agents Neural network Network of TLUs TLUs are thought to be simple models of biological neurons Connection weights Threshold value
Chapter 3: Neural Networks 3.2 Training Single TLUs Gradient Descent Widrow-Hoff Rule Generalized Delta Procedure 3.3 Neural Networks The Backpropagation Method Derivation of the Backpropagation Learning Rule 3.4 Generalization, Accuracy, and Overfitting
Overfitting
Chapter 4: Machine Evolution Genetic Algorithm Concept Genetic Programming How to define genetic operations and fitness function
Chapter 5: State Machines Concept of State Machine vs Stimulus-Response Agents
Chapter 6: Robot Vision Averaging Edge enhancement
Chapter 8: Uninformed Search Search Space Graphs Depth-First Search Breadth-First Search Iterative Deepening
Chapter 9 : Heuristic Search Using Evaluation Functions A General Graph-Searching Algorithm Algorithm A* Admissibility Consistency Iterative-Deepening A* Heuristic Functions and Search Efficiency
Chapter 10 Planning, Acting, and Learning X The Sense/Plan/Act Cycle X Approximate Search Learning Heuristic Functions Rewards Instead of Goals
Chapter 11: Not required
Chapter 12: Adversarial Search Minimax Procedure The Alpha-Beta Procedure Games of Chance Learning Evaluation Functions Not required
Chapter 13, 14, 15, 16 The Propositional Calculus Resolution in the Propositional Calculus The Predicate Calculus Resolution in the Predicate Calculus
Chapter 17: Knowledge-Based Systems Confronting the Real World Reasoning Using Horn Clauses Maintenance in Dynamic Knowledge Bases Rule-Based Expert Systems Rule Learning
Rule Extraction
Initialize Generic Separate-and-conquer algorithm (GSCA) Initialize Initialize empty set of rules repeat the outer loop adds rules until covers all (or most) of the positive instances Repeat the inner loop adds atoms to until covers only (or mainly) positive instances Select an atom to add to . This is a nondeterministic choice point that can be used for backtracking. Until covers only (or mainly) positive instances in We add the rule to the set of rules. (the positive instances in covered by ) Until covers all (or most) of positive instance in
Chapter 18. Representing commonsense Knowledge Not Required
Chapter 19. Reasoning with Uncertain Information Probabilistic Inference Bayes Networks Patterns of Inference in Bayes Networks Uncertain Evidence D-Separation Probabilistic Inference in Polytrees
Example
The rest chapters are not required