University of Sussex
Browse
isal_a_00196.pdf (1.94 MB)

Evolving recurrent neural network controllers by incremental fitness shaping

Download (1.94 MB)
Version 2 2023-06-12, 09:07
Version 1 2023-06-09, 17:59
conference contribution
posted on 2023-06-12, 09:07 authored by Kaan Akinci, Andy PhilippidesAndy Philippides
Time varying artificial neural networks are commonly used for dynamic problems such as games controllers and robotics as they give the controller a memory of what occurred in previous states which is important as actions in previous states can influence the final success of the agent. Because of this temporal dependence, methods such as back-propagation can be difficult to use to optimise network parameters and so genetic algorithms (GAs) are often used instead. While recurrent neural networks (RNNs) are a common network used with GAs, long short term memory (LSTM) networks have had less attention. Since, LSTM networks have a wide range of temporal dynamics, in this paper, we evolve an LSTM network as a controller for a lunar lander task with two evolutionary algorithms: a steady state GA (SSGA) and an evolutionary strategy (ES). Due to the presence of a large local optima in the fitness space, we implemented an incremental fitness scheme to both evolutionary algorithms. We also compare the behaviour and evolutionary progress of the LSTM with the behaviour of an RNN evolved via NEAT and ES with the same fitness function. LSTMs proved themselves to be evolvable on such tasks, though the SSGA solution was outperformed by the RNN. However, despite using an incremental scheme, the ES developed solutions far better than both showing that ES can be used both for incremental fitness and for LSTMs and RNNs on dynamic tasks.

History

Publication status

  • Published

File Version

  • Published version

Journal

Proceedings of the ALIFE Conference 2019 (ALIFE 2019): A Hybrid of the European Conference on Artificial Life (ECAL) and the International Conference on the Synthesis and Simulation of Living Systems (ALIFE)

Publisher

The MIT Press

Page range

416-423

Event name

ALIFE 2019

Event location

Newcastle upon Tyne, UK

Event type

conference

Event date

29 July - 2nd August, 2019

Department affiliated with

  • Informatics Publications

Research groups affiliated with

  • Centre for Computational Neuroscience and Robotics Publications

Full text available

  • Yes

Peer reviewed?

  • Yes

Editors

Harold Fellermann, Angel Goñi-Moreno, Jaume Bacardit, Rudolf Marcel Füchslin

Legacy Posted Date

2019-06-05

First Open Access (FOA) Date

2019-06-07

First Compliant Deposit (FCD) Date

2019-06-04

Usage metrics

    University of Sussex (Publications)

    Categories

    No categories selected

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC