This repository contains files pertaining to Assignment 02 of the course on the 'Foundations of Intelligent and Learning Agents' (CS 747) at IIT Bombay. The problem statement involves the implementation of the value iteration, linear programming, and Howard's policy iteration algorithms as they are, as well as using them to solve mazes with arbitrary start and end points.
A detailed report on the techniques utilized throughout the assignment has been provided about the methods used for the implementation and comparison of the algorithms.
submission contains the main code for optimal policy and value function evaluation, the encoder for translating the maze into an MDP, and the decoder for realizing a human-readable solution from the optimal policy, as well as references, planner verification and maze verification scripts, a grid visualization script, a custom MDP generation script and plots of both unsolved and solved mazes.