Low-Rank Tensor Regularized Graph Fuzzy Learning for Multi-View Data Processing

 Multi-view data processing is an effective tool to differentiate the  levels of consumers on electronics. Recently, the graph based multi-view  clustering methods have attracted widespread attention because they can  obtain the relationships of multi-view data points efficiently.  However, there exist several shortcomings on most existing graph based  clustering methods. Firstly, the mostly adopted Euclidean distance can  not extract the nonlinear manifold structure. Secondly, graph based  methods are mainly hard clustering methods, which means that each data  point belongs to only the one cluster exactly. Thirdly, the  high-dimension information between multiple views are not taken into  account. Thus, a low-rank tensor regularized graph fuzzy learning  (LRTGFL) method for multi-view data processing is proposed. In LRTGFL,  Jensen-Shannon divergence is adopted to replace the Euclidean distance  for obtaining more completely nonlinear structures. In addition, fuzzy  learning is adopted to make graph clustering be a soft clustering  method. Furthermore, a tensor nuclear norm based on the tensor singular  value decomposition (t-SVD) is adopted to take advantage of the  high-dimension information. Then, alternating direction method of  multipliers (ADMM) is adopted to solve the LRTGFL model. Finally, the  effectiveness and superiority of LRTGFL are demonstrated by comparing  with various state-of-the-art algorithms on eight real-world datasets. No description supplied