VSFCM: A Novel Viewpoint-Driven Subspace Fuzzy C-Means Algorithm

Nowadays, most fuzzy clustering algorithms are sensitive to the initialization results of and have a weak ability handle high-dimensional data. To solve these problems, we developed viewpoint-driven subspace c-means (VSFCM) algorithm. Firstly, propose new cut-off distance. Based on this, establish distance-induced (CDCI) method use it as strategy for cluster center viewpoint selection. Secondly, by taking obtained CDCI entry point knowledge, driven knowledge data is formed. upon these, put forward VSFCM algorithm combined with viewpoints, separation terms, feature weights. Moreover, compared symmetric weights other algorithms, exhibit significant asymmetry. That is, they assign greater features that contribute more, which validated artificial dataset DATA2 in experimental section. The multiple advanced three types datasets validate proposed has best performance five indicators. It demonstrated more effective, weight allocation consistent asymmetry data, can achieve better convergence speed while displaying efficiency.