GoT: a Growing Tree Model for Clustering Ensemble

The clustering ensemble technique that integrates multiple results can improve the accuracy and robustness of final clustering. In many algorithms, co-association matrix (CA matrix), which reflects frequency any two samples being partitioned into same cluster, plays an important role. However, generally, CA is highly sparse with low value density, may limit performance algorithm based on it. To handle these issues, in this paper, we propose a growing tree model (GoT). model, firstly refined by shortest path so its sparsity will be mitigated. Then, set representative prototype examples discovered. Finally, to density matrix, prototypes gradually connect their neighborhood, likes trees up. rationality discovered illustrated theoretical analysis experimental analysis. working mechanism GoT visually shown synthetic data sets. Experimental analyses eight UCI sets image show outperforms nine algorithms.