Bi-weighted ensemble via HMM-based approaches for temporal data clustering

作者：

Highlights：

• An effective solution is designed for the initialization problem based on the clustering ensemble technique.

• The performance of clustering ensemble is significantly improved via the proposed bi-weighting scheme.

• The intrinsic number of clusters can be automatically determined by applying agglomerative clustering in association with DSPA during the consensus process.

• Extensive experiments show that our proposed algorithm outperforms all the existing representative benchmarks.

摘要

•An effective solution is designed for the initialization problem based on the clustering ensemble technique.•The performance of clustering ensemble is significantly improved via the proposed bi-weighting scheme.•The intrinsic number of clusters can be automatically determined by applying agglomerative clustering in association with DSPA during the consensus process.•Extensive experiments show that our proposed algorithm outperforms all the existing representative benchmarks.

论文关键词：Data clustering,Ensemble learning,Hidden Markov Model,Model selection

论文评审过程：Received 22 November 2016, Revised 18 October 2017, Accepted 18 November 2017, Available online 21 November 2017, Version of Record 27 November 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.11.022