A New SSOPMV Learning for Matrix Data Sets

2018 2nd annual International Conference on Cloud Technology and Communication Engineering

计算机科学;无线电电子学

Zhu, Changming^1 ; Mei, Chengjiu^1 ; Zhou, Rigui^1 ; Wei, Lai^1 ; Zhang, Xiafen^1 ; Yao, Min^1

College of Information Engineering, Shanghai Maritime University, No. 1550, LinGang Ave., Shanghai, China^1

关键词: Learning machines;    Multi-view datum;    Multi-view learning;    Multi-views;    Process matrix;    Real-world;    Semi-supervised;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/466/1/012111/pdf
DOI  :  10.1088/1757-899X/466/1/012111

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

In real-world applications, most multi-view data sets are semi-supervised and large-scale. In order to process these data sets, scholars have developed semi-supervise done-pass multi-view learning (SSOPMV). While SSOPMV cannot process matrix data sets. Thus this manuscript extends the model of SSOPMV to matrix version and the new learning machine is named matrix-instance-based SSOPMV, i.e. (MSSOPMV). Related experiments validate that MSSOPMV can process multi-view, semi-supervised, large-scale, and matrix data sets well.