PECST522 · S5 CSE · 2026-27

Artificial Intelligence.

Four modules — intelligent agents, problem solving by search, knowledge representation & reasoning, reinforcement learning. Slides as PDF. Hands-on: Bench Answer Blitz, AI-assisted labs on NVIDIA NIM.

Modules / 01

MODULE 1LIVE
Intelligent Agents
Agents, environments, PEAS descriptions, four agent architectures. Bench Answer Blitz portal for M1 top-frequency questions.
PDF · Portal
MODULE 2LIVE
Problem Solving by Search
Uninformed & informed search, BFS/DFS/UCS/A*, heuristics, adversarial search. Companion game: N-Queens Solver.
PDF · Game
MODULE 3LIVE
Knowledge Rep & Reasoning
Propositional & first-order logic, resolution, forward/backward chaining, Bayesian networks. Companion game: Wumpus World.
PDF · Game
MODULE 4LIVE
Reinforcement Learning
MDPs, value iteration, Q-learning, policy gradients. NVIDIA NIM API lab next week.
PDF · Lab

Tools for this course / 02

PORTAL
Bench Answer Blitz
In-class question portal, benches answer 6 top-frequency M1 questions, admin panel for review.
Open portal
GAME
Agent Program Playground
Program your own rules, compare all four architectures head-to-head.
Play
GAME
Wumpus World
Classic KR&R exemplar — a knowledge-based agent reasons about breeze/stench/glitter to hunt gold and avoid pits. Companion to Module 3.
Play
GAME
Logic Lab
Build propositional-logic derivations step by step — reasoning patterns, CNF, resolution, forward/backward chaining. Every line is checked instantly; Unstuck hints and an optional AI tutor. Companion to Module 3.
Play

Class list / 03