A Multi-View Clustering Algorithm for Mixed Numeric and Categorical Data

Clustering data with both numeric and categorical attributes is of great importance as such are ubiquitous in real-world problems. Multi-view learning approaches have proven to be more effective having better generalisation ability compared single-view many However, most the existing clustering algorithms developed for mixed single-view. In this research, we propose a novel multi-view algorithm based on k-prototypes (which term K-Prototypes) data. To best our knowledge, proposed K-Prototypes first version well-known algorithm. cluster over multiple views, present representation prototype centres scenario also devise formulas updating each view. Then concept consensus output final result. Finally, carried out series experiments four benchmark datasets assess performance clustering. Experimental results show that outperforms seven state-of-the-art cases.