Novel Reinforcement Learning Research Platform for Role-Playing Games

The latest achievements in the field of reinforcement learning have encouraged the development of vision-based learning methods that compete with human-provided results obtained on various games and training environments. Convolutional neural networks together with Q-learning-based approaches have m...

Full description

Saved in:
Bibliographic Details
Published in:Mathematics (Basel) 2022-11, Vol.10 (22), p.4363
Main Authors: Csereoka, Petra, Roman, Bogdan-Ionuţ, Micea , Mihai Victor, Popa, Călin-Adrian
Format: Article
Language:English
Subjects:
Artificial neural networks
Collaboration
Dark Souls III
Deep learning
Deep Q-Network
Deep Recurrent Q-Network
Games
Machine learning
Neural networks
Players
Role playing
Source code
Training
video games
visual-based reinforcement learning
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The latest achievements in the field of reinforcement learning have encouraged the development of vision-based learning methods that compete with human-provided results obtained on various games and training environments. Convolutional neural networks together with Q-learning-based approaches have managed to solve and outperform human players in environments such as Atari 2600, Doom or StarCraft II, but the niche of 3D realistic games with a high degree of freedom of movement and rich graphics remains unexplored, despite having the highest resemblance to real-world situations. In this paper, we propose a novel testbed to push the limits of deep learning methods, namely an OpenAI Gym-like environment based on Dark Souls III, a notoriously difficult role-playing game, where even human players have reportedly struggled. We explore two types of architectures, Deep Q-Network and Deep Recurrent Q-Network, providing the results of a first incursion into this new problem class. The source code for the training environment and baselines is made available.
ISSN:2227-7390
2227-7390
DOI:10.3390/math10224363