A new unconstrained global optimization method based on clustering and parabolic approximation
详细信息 查看全文
- 作者：Ihsan Pencea ; ihsanpence@mehmetakif.edu.tr" class="auth_mail" title="E-mail the corresponding author ; ihsan_ce@hotmail.com" class="auth_mail" title="E-mail the corresponding author ; Melike Siseci Cesmelib ; melikesiseci@mehmetakif.edu.tr" class="auth_mail" title="E-mail the corresponding author ; Fatih Ahmet Senelc ; fatihsenel@sdu.edu.tr" class="auth_mail" title="E-mail the corresponding author ; Bayram Cetislic ; bayramcetisli@sdu.edu.tr" class="auth_mail" title="E-mail the corresponding author
- 关键词：Unconstrained global optimization ; Fuzzy c-means ; Clustering ; Least squares estimation ; Parabolic approximation
- 刊名：Expert Systems with Applications
- 出版年：2016
- 出版时间：15 August 2016
- 年：2016
- 卷：55
- 期：Complete
- 页码：493-507
- 全文大小：2699 K
文摘
- •
ara0001">Proposed method uses clustering and parabolic approximation.
- •
ara0002">Clustering period provides jumping to better points using with populations.
- •
ara0003">The vertex of parabola is much close to the local minima than obtained cluster center.
- •
ara0004">We assume that the best cluster center denotes the position of global optimum.
- •
ara0005">The proposed method is simple, faster and very successful to gain the global optimum.