<conference paper>
Optimization Method Using Information on Fitness Landscape

Creator
Language
Publisher
Date
First Page
Last Page
Conference
Publication Type
Access Rights
Abstract We propose a stochastic optimization method that uses fitness landscape information. This method (1) approximates the landscape by linearly combining multiple Gaussian functions and (2) generates new ...searching points by regarding the Gaussian functions as probabilistic density functions. Since the method can grasp the landscape roughly and globally, it is expected that better searching area is quickly found than other optimization methods that do not use whole landscape information. We apply the method to benchmark functions, evaluate its convergence characteristics, and discuss its advantage and disadvantage.show more
Table of Contents 1 はじめに
2 提案手法の概要
3 ベンチマーク関数を用いた評価
4 全体の考察
5 まとめ

Hide fulltext details.

pdf DomesticConf080 pdf 1.50 MB 133  

Details

Record ID
Created Date 2021.07.16
Modified Date 2021.08.02

People who viewed this item also viewed