论文部分内容阅读
D-S证据组合规则在处理高冲突信息时会得出与直觉相反的结论,这一直是D-S理论研究的热点.与相关理论优势互补是克服证据理论固有缺陷的有效方法之一.基于对最大熵原理和证据理论的研究,定义了辨识框架上的基本最大熵置信分配函数,并与经典的D-S组合规则及其改进方法相结合,给出了相关推理公式及基于置信最大熵模型.理论分析和实验表明,最大熵新证据的加入使非单焦元的基本置信赋值按比例重新分配给了单焦元,很好地处理了高冲突信息.
DS evidence combination rules in dealing with high conflict information will come to an opposite conclusion and intuition, which has been a hot topic in DS theory.Completing with the related theories is one of the effective ways to overcome the inherent defects of evidence theory.Based on the principle of maximum entropy And evidence theory, we define the basic maximum entropy confidence assignment function on the identification framework, and combine with the classical DS combination rule and its improved method, and give the relevant reasoning formula and confidence maximum entropy model. Theoretical analysis and experiment It is shown that the addition of new evidence of maximum entropy reassigns the basic confidence assignment of non-single-focal to the single-focal in proportion and deals well with the high-conflict information.