Adaptive Agents in 1v1 Snake Game with Dynamic Environment

Publikation: Kapitel i bog/rapport/konference-proceedingKonferencebidrag i proceedingsForskningpeer review

4 Downloads (Pure)

Abstract

This paper delves into the adaptability of Proximal Policy Optimization (PPO)-trained agents within dynamic environments. Typically, an agent is trained within a specific environment, learning to maximise reward acquisition and to navigate it effectively. However, alterations to this environment can lead to performance deficiencies. Existing research does not fully elucidate how the training of agents influences their adaptability in different environments and which parameters significantly impact this. This study aims to fill this gap, contributing to the creation of more versatile intelligent agents. The objective of this study is to explore how training agents in various environments affects their adaptability when introduced to unfamiliar environments. To this end, 36 models were trained using 36 different configurations to play a one-versus-one (1v1) Snake game. These models were subsequently compared against each configuration to measure their adaptability. The results reveal that map size substantially affect the adaptability of agents in different environments. Interestingly, the results showed that the most adaptive agents were not those trained on the most expansive and complex environment, but rather the simplest.

OriginalsprogEngelsk
Titel2024 IEEE Conference on Games (CoG)
Antal sider4
ForlagIEEE Press
Publikationsdato2024
ISBN (Elektronisk)9798350350678
DOI
StatusUdgivet - 2024
Begivenhed6th Annual IEEE Conference on Games, CoG 2024 - Milan, Italien
Varighed: 5. aug. 20248. aug. 2024

Konference

Konference6th Annual IEEE Conference on Games, CoG 2024
Land/OmrådeItalien
ByMilan
Periode05/08/202408/08/2024
SponsorHotel Principe di Savoia Milano - Dorchester Collection, IEEE, IEEE Computational Intelligence Society, Politecnico Milano, University of Milan, Universitas Studiorum Mediolanensis
NavnProceedings - IEEE Conference on Games
ISSN2325-4270

Bibliografisk note

Publisher Copyright:
© 2024 IEEE.

Fingeraftryk

Dyk ned i forskningsemnerne om 'Adaptive Agents in 1v1 Snake Game with Dynamic Environment'. Sammen danner de et unikt fingeraftryk.

Citationsformater