Evolutionary Multi-objective Clustering Over Multiple Conflicting Data Views

Multi-view data analysis provides an effective means to integrate the distinct information sources which are inherent many applications. Data clustering in a multi-view setting specifically aims identify most appropriate grouping for collection of entities, where those entities (or their relationships) can be described from multiple perspectives. Leveraging recent advances multi-objective clustering, we propose new evolutionary method tackle this challenge. Designed around flexible and unbiased solution representation, together with strategies based on minimum spanning tree neighborhood relations, our algorithm optimizes objectives simultaneously effectively explore space candidate trade-offs between views. Through series experiments, investigate suitability proposal context bioinformatics application, plausible protein structures, diverse set synthetic problems. The specific case two views is considered paper. evaluation respect variety reference approaches demonstrates effectiveness discovering high-quality partitions multiview setting. Robustness against unreliable ability automatically determine number clusters, additional advantages evidenced by results obtained.