AI, Monsoon 2026

Session

Topic

Topic details

Readings

Slides

Exercises

1

Introduction

What is AI?, goals of AI, AI applications, foundations of AI, history of AI

1

2

Intelligent Agents

Agents and Environment, Nature of Environment, Structure of Agent, Rationality, Performance Measures

2

3, 4

Problem Solving Agents and Search

Problem Solving Agents, Best-First Search, Uninformed Search: Breadth First Search, Depth First Search, Iterative Deepening DFS

3.1-3.4

5

Informed Search

Heuristic Functions, Greedy Best First search, A* Search

3.5-3.6

6, 7

Constraint Satisfaction Problems

CSP, Inference in CSPs, Backtracking and Local Search for CSPs

6.1-6.5

8

Propositional Logic

Introduction to Logic, Propositional Logic: syntax and semantics, model checking, resolution sketch, Knowledge-based Agents

7.1-7.7

9

First-Order Logic

FOL syntax and semantics, using FOL, knowledge engineering

8.1-8.4

10, 11

Probability Review

Basics, Conditional Probability, Bayes' Theorem, Naive Bayes Classifier

12

12, 13

Probabilistic Reasoning

Bayesian Networks, Representation, Conditional Independence, Exact Inference (Enumeration and Variable Elimination)

13.1-13.5

Mid-Semester Exam

[Sept 19-27, 2026] Topics covered till date

14

Probabilistic Reasoning over Time

Time and Uncertainty, Markov Chains, Hidden Markov Models (HMMs), filtering, Viterbi sketch

14.1-14.5

15

Utility Theory and Decision Networks

Utility Theory, Utility Functions, Decision Networks, Value of Perfect Information

16.1-16.6

16

Sequential Decision Making - I

Markov Decision Processes (MDPs): States, Actions, Rewards, Transition Function, Policy

17.1

17

Sequential Decision Making - II

Bellman Equations, MDP Algorithms: Value Iteration, Policy Iteration

17.2-3

18

Reinforcement Learning

Introduction to Reinforcement Learning (RL), Agent-Environment Interaction, Learning from Rewards

22

19

Machine Learning and Neural Networks

Supervised learning setup, linear and logistic regression, perceptron, multi-layer perceptron, activation functions, backpropagation intuition. Brief linear algebra refresher.

19.1-19.6, 21.1-21.2

20

Deep Learning

SGD, regularization, CNN sketch. Function approximation as a unifying lens (incl. for RL value functions).

21.3-21.6

21, 22

Transformers and Large Language Models

NLP, word embeddings, tokenization, attention as soft lookup, transformer block, pre-training vs. fine-tuning, in-context learning, prompting, retrieval-augmented generation (RAG). Connecting LLMs back to agents and search (LLM-as-controller, tool use).

24

23

AI Ethics, Safety and Alignment

Bias, hallucination, evaluation, Reinforcement Learning From Human Feedback (RLHF) at a high level, dual-use concerns, societal impact, regulation.

27

24

Project presentations

25

Epilogue / Capstone Synthesis

Tying the agentic-AI loop (perceive → reason → act) through every unit. Summarisation, Quiz.