Fuzzy-Rough Intrigued Harmonic Discrepancy Clustering

Fuzzy clustering decomposes data into clusters using partial memberships by exploring the cluster structure information, which demonstrates comparable performance for knowledge exploitation under circumstance of information incompleteness. In general, this scheme considers objects to centroids and applies with spherical distribution. addition, noises outliers may significantly influence process; a common mitigation measure is application separate noise processing algorithms, but usually introduces multiple parameters are challenging be determined different types. This paper proposes new fuzzy-rough intrigued harmonic discrepancy (HDC) algorithm noting that sets offer higher degree uncertainty modelling both vagueness imprecision present in real-valued datasets. The HDC implemented introducing novel concept discrepancy, effectively indicates dissimilarity between instance foreign their distributions fully considered. proposed thus featured powerful ability on complex distribution leading enhanced performance, particularly noisy datasets, without use explicit handling parameters. experimental results confirm effectiveness HDC, generally outperforms popular representative algorithms synthetic benchmark demonstrating superiority algorithm.