Ambiguity Multi-attribute Decisions with Evidential Chains-based Fusion Reasoning Method

摘要

To learn the distribution of outcomes from previously ambiguity information for multi-attribute decision making, a novel associative Evidential Chains (ECs)-based fusion reasoning method was proposed. The convex combination of probabilistic beliefs from multiple refined ECs were induced by the proposed model based on multiple criteria linear programming. Its solution for the query case has varied flexibility with the similarity matrix derived from the historical ECs. Results of the applications with the benchmark cardiac diagnostic data sets verify that the proposed method is effective and interpretable.