Similarity Measures between Connection Numbers of Set Pair Analysis

The Set Pair Analysis (SPA) is a new system analysis approach and uncertainty theory. The similarity measure between connection numbers is the key to applications of SPA in multi-attribute decision-making, pattern recognition, artificial intelligent. However, it is difficult to accurately depict similarity degree between connection numbers. The distance between connection numbers, a group of checking criterions and the similarity degree functions of connection numbers in SPA are presented in this paper to measure the similarity between connection numbers, and the rationality of such measurement is also explained by the well-designed criterions. The result shows the effectiveness of the proposed similarity measures.