Implementation of reinforcement learning algorithms

Algorithms included:

Temporal-Difference (TD) learning algorithms: Q-learning, Sarsa

n-steps Boostrapping: n-steps Sarsa

Linear function approximation: Semi-gradient Sarsa with tile encoding

Deep Reinforcement Learning: DQN

All algorithms are tested with MountainCar-v0 environment in experiments/. Use plot_learning.py to plot the results.