EE 556 Final Project
In this project, we have implemented deep reinforcement learning using DQN and DDQN agents with and without Priortized Experience replay on several agents. In second part of project, we have implemented Quantile regression Distribution learning and observed the rewards gained by environments
Convolution Neural Network:
DQN, DDQN, PER results for CartPole-v0 and LunarLander-v2
Quantile Regression Results
CartPole and LunarLander
Super Mario Bros
