Reinforcement Learning
Contains solution to assignment problems of the course "CS404 : Reinforcement Learning" offered by Dr. Prabuchandran K J at IIT Dharwad.
-
Assignment 2 : Bandit Algorithms
- Implemented epsilon-greedy, variable epsilon-greedy, Softmax, Upper Confidence Bound (UCB) and Thompson sampling algorithms for Bernoulli and Normal reward setting
-
Assignment 3 : Value Based Methods
- A classical maze problem was considered and policy iteration and value iteration were used to solve the problem.