CS747-FILA

Author: Manas Vashistha

---

• Assignment for the course CS747 Foundations of Intelligent and Learning Agents under Prof Shivaram Kalyanakrishnan.

Assignment 1

Estimating average regrets for Multiarmed Bandit Instances using $\epsilon$-greedy, UCB, KL-UCB and Thompson Sampling.

Assignment 2

• Computing the optimal value functions for MDPs using Value iteration, Linear Programming and Policy improvement.
• Formulating a maze as an mdp to find the shortest path from a starting state to an end state.

Assignment 3

Windy Gridworld task.