Bayesian Ensembles for Exploration in Deep Reinforcement Learning; Code underlying the dissertation "Bayesian Model-Free Deep Reinforcement Learning"

Code underlying the dissertation "Bayesian Model-Free Deep Reinforcement Learning". This repository contains code for the chapter "Bayesian Ensembles for Exploration in Deep Q-Learning". The code is for experiments which demonstrate how ensemble-based Bayesian methods can improve exploration efficiency in reinforcement learning, with a focus on the DQN architecture. The code can be used to reproduce the results and be modified by researchers to be applied to other benchmark problems. The code has no direct applications outside of reinforcement learning research.