Towards improvement of SUNA in multiplexers with preliminary results of simple logic gate neuron variation

研究成果: Chapter in Book/Report/Conference proceedingConference contribution

抄録

Spectrum-Diverse Neuroevolution with Unified Neural Models (SUNA) has been shown to be a successful alternative to the algorithm NeuroEvolution of Augmenting Topologies (NEAT). Requiring less parameters than NEAT yet possessing a more unified representation power and effective spectrum-based diversity preservation, SUNA outperformed NEAT on most of the problems to be experimented. However, we think a simple improvement approach can be made to improve SUNA's efficiency in the strategic decision-making problem tested by the model itself, i.e. the multiplexer problem. In the proposed method, we try to incorporate the idea of logical gates to the hidden neurons in the model, suggesting it the solutions that solve the problem in the real world in the form of neurons. It is shown that with the simple logic gates neuron variations, SUNA can be slightly enhanced to resolve the multiplexer problem.

本文言語英語
ホスト出版物のタイトルGECCO 2020 Companion - Proceedings of the 2020 Genetic and Evolutionary Computation Conference Companion
出版社Association for Computing Machinery, Inc
ページ53-54
ページ数2
ISBN(電子版)9781450371278
DOI
出版ステータス出版済み - 7 8 2020
イベント2020 Genetic and Evolutionary Computation Conference, GECCO 2020 - Cancun, メキシコ
継続期間: 7 8 20207 12 2020

出版物シリーズ

名前GECCO 2020 Companion - Proceedings of the 2020 Genetic and Evolutionary Computation Conference Companion

会議

会議2020 Genetic and Evolutionary Computation Conference, GECCO 2020
国/地域メキシコ
CityCancun
Period7/8/207/12/20

All Science Journal Classification (ASJC) codes

  • 計算数学

フィンガープリント

「Towards improvement of SUNA in multiplexers with preliminary results of simple logic gate neuron variation」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル