Agricultural Journal
ISSN: Online 1994-4616ISSN: Print 1816-9155
Abstract
Building an agent to play games might be done in several ways, like mini-max, Monte Carlo tree search, deep learning or it could be a combination of two or three technics or even more, like the popular chess computer engine deep blue or alpha go. This document built an agent that plays and balances cart pole game, the agent used deep reinforcement learning-specifically Q-learning algorithm to build neural network that greedy to maximize the reward function and balance the pole for the longest period at the end we got an agent that outperform the random and human agent.
How to cite this article:
Abdallah Al-Zu’bi and Ahmad Al-Qerem. Deep Reinforcement Learning Applied to Cart Pole Game.
DOI: https://doi.org/10.36478/aj.2020.137.142
URL: https://www.makhillpublications.co/view-article/1816-9155/aj.2020.137.142