Improving consensus clustering with noise-induced ensemble generation
作者：
Highlights：
• Attribute-noise-induced ensemble generation is proposed for consensus clustering.
• It exploits different density functions: Gaussian, Exponential and uniform random.
• It outperforms baseline & instance-wise models, 10 data sets & 4 validity indices.
• Data-dependent context for its application is given: appropriate noise type & ratio.
• This study also provides investigation of parameter settings and complexity.
摘要
•Attribute-noise-induced ensemble generation is proposed for consensus clustering.•It exploits different density functions: Gaussian, Exponential and uniform random.•It outperforms baseline & instance-wise models, 10 data sets & 4 validity indices.•Data-dependent context for its application is given: appropriate noise type & ratio.•This study also provides investigation of parameter settings and complexity.
论文关键词：Consensus clustering,Attribute noise,Ensemble generation
论文评审过程：Received 18 August 2019, Revised 14 December 2019, Accepted 14 December 2019, Available online 24 December 2019, Version of Record 30 December 2019.
论文官网地址：https://doi.org/10.1016/j.eswa.2019.113138
