Modul pro Reinforcment Learning pro Modeler neuronových sítí

Abstract

This master thesis focuses on extending the program Neural Net Modeler with Reinforcement Learning module. This thesis describes Reinforcement learning with and without deep learning. Furthermore, this thesis deals with problems of Reinforcement learning and their solutions. Further, I describe the parallelization of Reinforcement learning with neural networks. Using implemented solutions, experiments were performed and results compared.