XIV International Scientific Conference «Electronics and Applied Physics»

OPTIMIZING ALGORITHM, BASED ON REINFORCEMENT LEARNING AND NEURAL NETWORK

Not scheduled

20m

Faculty of Radio Physics, Electronics and Computer Systems

Oral Computer Technologies Computer Technologies

Speaker

Mr Oleksii Morozov (NSBC consulting company)

Abstract

A long-standing goal of artificial intelligence is an algorithm that learns, tabula rasa, superhuman proficiency in challenging domains. We propose the algorithm, based on neural network, capable to learn how to play any 2-player board game to high master level, using only simple rules of this game as an input. Here, we introduce an algorithm based solely on reinforcement learning, without any human data, guidance, or domain knowledge beyond game rules.