Home // DIGITAL 2022, Advances on Societal Digital Transformation // View article

Toward Comparing Knowledge Acquisition in DeepRL Models

Authors:
Anthony Marchiafava
Atriya Sen

Keywords: Artificial Intelligence, Deep Reinforcement Learning, Reinforcement Learning, Deep Learning, Explainable AI

Abstract:
In order to better exploit Deep Reinforcement Learning (DeepRL) systems such as DeepMind’s Alpha Go & Alpha Zero, it is desirable to understand how they acquire knowledge, and how human knowledge acquisition can contribute to or benefit from such an understanding. We analyze a series of DeepRL models trained to play the board game of chess in a human-like fashion, to study if these models acquire concepts differently from self-trained DeepRL models such as AlphaZero. Our preliminary results indicate that human chess players may acquire concepts very similarly to self-trained models. We further discuss some of the potential consequences of such an outcome.

Pages: 13 to 16

Copyright: Copyright (c) IARIA, 2022

Publication date: November 13, 2022

Published in: conference

ISBN: 978-1-68558-014-8

Location: Valencia, Spain

Dates: from November 13, 2022 to November 17, 2022