2017
DOI: 10.1109/tciaig.2015.2494596
|View full text |Cite
|
Sign up to set email alerts
|

Neuroevolution in Games: State of the Art and Open Challenges

Abstract: This paper surveys research on applying neuroevolution (NE) to games. In neuroevolution, artificial neural networks are trained through evolutionary algorithms, taking inspiration from the way biological brains evolved. We analyse the application of NE in games along five different axes, which are the role NE is chosen to play in a game, the different types of neural networks used, the way these networks are evolved, how the fitness is determined and what type of input the network receives.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
56
0
2

Year Published

2019
2019
2022
2022

Publication Types

Select...
5
2
1

Relationship

2
6

Authors

Journals

citations
Cited by 123 publications
(58 citation statements)
references
References 119 publications
(197 reference statements)
0
56
0
2
Order By: Relevance
“…Interestingly, separate evolutionary runs often follow similar performance curves (which we also observed in other experiments). This behaviour appears very different from the training of networks with orders of magnitude fewer parameters traditionally studied in neuroevolution, which often have a higher variance across runs [7,32,39]. Analyzing this phenomenon in more detail is an interesting future research direction that we aim to investigate.…”
Section: Methodsmentioning
confidence: 92%
“…Interestingly, separate evolutionary runs often follow similar performance curves (which we also observed in other experiments). This behaviour appears very different from the training of networks with orders of magnitude fewer parameters traditionally studied in neuroevolution, which often have a higher variance across runs [7,32,39]. Analyzing this phenomenon in more detail is an interesting future research direction that we aim to investigate.…”
Section: Methodsmentioning
confidence: 92%
“…Because of their generality, NE approaches have been applied extensively to different types of video games. For a complete overview of this field, we refer the interested reader to our NE survey paper [115].…”
Section: ) Evolutionary Approachesmentioning
confidence: 99%
“…DQN [97] was very influential as an algorithm that uses gradient-based deep learning for pixel-based video game playing and was originally applied to the Atari benchmark. Note that earlier approaches exist but with less success such as [109], and successful gradient-free methods [115]. Double DQN [155] and Dueling DQN [161] are early extensions that use multiple networks to improve estimations.…”
Section: Historical Overview Of Deep Learning In Gamesmentioning
confidence: 99%
“…De acordo com a pesquisa de literatura feita em [4], na maior parte dos casos a neuroevoluçãoé usada para evoluir um agente na tarefa de jogar um jogo.É importante notar que existem diferentes tipos de desafios ao se evoluir um agente, e isto pode variar de acordo com o jogo. Alguns destes desafios são: avaliação de estado/ação, geração de conteúdo, seleção de estratégias e seleção de ação direta [4], sendo esteúltimo o tipo de desafio do jogo que serviu de ambiente para os experimentos realizados neste artigo.…”
Section: Neuroevoluçãounclassified
“…Alguns destes desafios são: avaliação de estado/ação, geração de conteúdo, seleção de estratégias e seleção de ação direta [4], sendo esteúltimo o tipo de desafio do jogo que serviu de ambiente para os experimentos realizados neste artigo.…”
Section: Neuroevoluçãounclassified