Acquisition of all credible rules and knowledge reductions in incomplete ordered interval-valued decision system
1.School of Mechanical Engineering, Nanjing University of Science and Technology, Nanjing, Jiangsu 210094, China; 2.Unit 73106 of PLA, Huaian, Jiangsu 223300, China; 3.School of Computer Science and Telecommunication Engineering, Jiangsu University, Zhenjiang, Jiangsu 212013, China
Abstract：The approaches for the acquisition of all credible rules in incomplete ordered interval-valued decision systems were investigated. In the systems, not only unknown values were included, but also all attributes were considered as criterions. The concepts of missing and filling interval data were proposed. The character of interval data was deeply investigated, and a concept of interval-segment data was proposed to construct a new interval-segment data ordered decision system. The concepts of the increasing and decreasing descriptors were proposed based on interval-segment data to induce all of the credible rules and to provide the specific algorithm steps. The verified results by an illustrative example show that the proposed method is effective.
2014, Vol. 35 
