Artificial Intelligence Notes | B.Sc. (Computer Science) Semester 5 | Mumbai University | munotes

Contents

Module I Intelligent agents, problem solving and search, knowledge representation and logical reasoning, and reasoning under uncertainty

1. What Artificial Intelligence Is 1
2. Acting Rationally, or Thinking Like a Human 6
3. The Agent and Its Environment 10
4. Types of Environment 15
5. The Simple Reflex Agent 20
6. The Model-Based Reflex Agent 24
7. The Goal-Based Agent 29
8. The Utility-Based Agent 34
9. The Learning Agent 39
10. Formulating a Problem for Search 45
11. The Search Tree, and How an Algorithm Is Judged 51
12. Breadth First Search 57
13. Depth First Search 62
14. Uniform Cost Search 67
15. Iterative Deepening Search 73
16. Heuristics: Estimating What Is Left To Do 78
17. Greedy Best First Search 85
18. A* Search 90
19. Admissibility, and Why A* Is Optimal 96
20. Consistency, and A* on a Graph 102
21. Recursive Best First Search 108
22. Comparing Two Search Algorithms on One Problem 113
23. Games as Search 119
24. Minimax 124
25. Alpha-Beta Pruning 129
26. The Knowledge-Based Agent 136
27. Propositional Logic: Syntax and Meaning 142
28. Entailment, Models and Validity 147
29. Inference in Propositional Logic 152
30. Conjunctive Normal Form and Resolution 157
31. First-Order Logic 163
32. Inference in First-Order Logic: Unification and Chaining 168
33. Rule-Based Systems and Expert Systems 175
34. Planning and STRIPS 181
35. Fuzzy Logic and the Membership Function 187
36. Fuzzification, the Rule Base and Defuzzification 193
37. Why an Agent Needs Probability 199
38. The Joint Distribution and Inference by Enumeration 204
39. Bayes Theorem 209
40. Conditional Independence 215
41. Bayesian Networks 221
42. Inference in a Bayesian Network 227

Module II Machine learning, supervised models, probabilistic and latent variable models, unsupervised and reinforcement learning, and responsible AI

1. What Machine Learning Is 234
2. Supervised Learning 240
3. Unsupervised Learning 246
4. Reinforcement Learning: The Third Form 250
5. Parametric and Nonparametric Models 254
6. Bias and Variance 259
7. Overfitting and Underfitting 265
8. Regularization 271
9. Gradient Descent 277
10. Classification and Regression 284
11. k-Nearest Neighbours 290
12. k-NN for Regression, and What Limits the Method 296
13. Entropy and Information Gain 302
14. Decision Tree Learning 308
15. Reading, Drawing and Pruning a Decision Tree 314
16. The Naive Bayes Classifier 321
17. Support Vector Machines: The Margin 328
18. The Soft Margin and the Kernel 334
19. The Artificial Neuron and the Perceptron 342
20. The Multilayer Network and Backpropagation 349
21. What Deep Learning Is 357
22. Ensemble Methods, Bagging and the Random Forest 362
23. Boosting and AdaBoost 370
24. Evaluating a Model 378
25. The Statistical Learning Framework 387
26. Maximum Likelihood Estimation 395
27. Learning with Complete Data 404
28. Hidden Variables 412
29. The EM Algorithm 420
30. Hidden Markov Models 430
31. The Forward Algorithm and Viterbi 439
32. Clustering and k-Means 451
33. Hierarchical Clustering, and Judging a Clustering 462
34. Support, Confidence and Lift 475
35. The Apriori Algorithm 485
36. The Reinforcement Learning Framework 495
37. Markov Decision Processes 505
38. The Bellman Equations, Value Iteration and Policy Iteration 514
39. Q-Learning 525
40. Ethical Issues in AI Systems 537
41. Bias and Fairness in AI Models 543
42. Transparency and Explainability 553
43. Accountability and Human Oversight 559
44. Hallucination in Generative AI 564
45. Deepfakes and Misuse 575
46. Using an AI Library Responsibly 581
47. What the Examination Asks, and How to Answer It 587