Memetic differential evolution methods for clustering problems

作者：

Highlights：

• We define a new population based algorithm for the euclidean minimum sum of squares clustering problem.

• The new algorithm includes a matching phase to align different clustering solutions.

• The algorithm uses K-means as a local optimization routine.

• The numerical performance is excellent both in terms of quality and of computational requirements, when compared to state of the art.

摘要

•We define a new population based algorithm for the euclidean minimum sum of squares clustering problem.•The new algorithm includes a matching phase to align different clustering solutions.•The algorithm uses K-means as a local optimization routine.•The numerical performance is excellent both in terms of quality and of computational requirements, when compared to state of the art.

论文关键词：Global optimization,Clustering,Minimum sum-of-squares,Hybrid genetic algorithm,K-MEANS

论文评审过程：Received 12 June 2020, Revised 23 December 2020, Accepted 22 January 2021, Available online 27 January 2021, Version of Record 11 February 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.107849