A Multi-view Decision Model Based on CCA

Jie Chen, Shu Zhao, Yanping Zhang

Abstract: Three-way decision theory divides all samples into three regions: positive region, negative region and boundary region. A lack of detailed information may make a definite decision impossible for samples in boundary region. These samples may be further handled by using new information. In this paper, we propose a method Multi-View Decision Model based on constructive three-way decision theory. Multi-view Decision Model mines the global information of all samples for decision. All samples firstly are decided by MinCA, which builds the min covers for each class. Then samples in boundary region are classified using Multi-view information. Experiments have shown that in most cases, Multi-View Decision Model is beneficial for reducing boundary region and promoting classification precision.