Hranie počítačových hier pomocou umelej inteligencie

Abstract

The goal of the bachelor thesis is to implement the game Pong and then convert it to autonomous play using algorithms. In this thesis there are two implementations one using QL algorithm which uses Q-Table and the other using DQL algorithm which uses DNN. Agent will learn against an unbeatable player according to which it should be able to learn to copy his moves.

Description

Subject(s)

Reinforced Learning, Q-Learning, Deep Q-Learning, Pong

Citation