Research on Information Fusion Algorithm Based on Multi-Agent Group Decision Mechanism
Yachong Tian, Wei Jin, Tao Huang, Xianyi Cheng · Advances in transdisciplinary engineering · 2022
Aim at the problem of rough and low precision in the classifier fusion of sorting layer output, an information fusion algorithm based on multi-agent theory is presented. This algorithm proposes based on multi-agent as classifier combination. It applies the collating of individual agent are classified with Clustering Analysis method, according to the result of classification, multi-agent is classified into several groups and the weight coefficient is given to every agent. By building relative matrixes between sort set and sort making order, and classified sort set on relative matrixes, the method constructs an optimization decision making model. According to fused group consensus in rank making. After those are achieved, combination decision is made. According to experiments on speech emotion recognition database, the result of the experience indicates the model is simple and feasible.