A New Basic Probability Assignment Method for Multi-Source and High Conflict Information and its Application
Xin Guan, Jing Zhao, Hai Qiao Liu · Materials science forum · 2019
In the field of target detection, material processing, such as nondestructive testing, welding and joining, many sensors are used to acquire information from different sources. In order to solve the problem of low precision of fusion results caused by conflicts between multi-source information, a new multi-source high-conflict information basic probability assignment method without any prior knowledge is proposed. Three most basic classifiers and the random selection of multiple sets of data in the UCI international public database are selected for simulation experiments to verify the feasibility and effectiveness of the proposed new method.