一种快速分层递阶DSmT近似推理融合方法(A)

 本文提出了一种分层递阶的DSmT快速近似推理融合方法,该方法针对超幂集空间中仅单子焦元具有信度赋值的情况,利用二叉树或三叉树分组技术对其刚性分组,与此同时,对每个信息源对应的各个分组焦元进行信度赋值求和,以便实现细粒度超幂集空间向粗粒度超幂集空间映射.然后运用DSmT组合规则和比例冲突分配规则对粗化超幂集空间的两个信息源进行融合,保存该融合结果作为父子之间节点连接权值,然后对每个分组焦元信度赋值归一化处理,通过设定树的深度,来确定分层递阶的次数.最后通过从多个角度比较新、老方法,从而充分地验证了新方法的优越性.

A Fast Approximate Reasoning Method in Hierarchical DSmT(A)

A kind of fast approximate reasoning method in hierarchical DSmT is proposed.This method is only fit for the case that there are only singleton focal elements with assignments in hyper-power set.These focal elements in hyper-power set are forced to group through bintree or tritree technologies.At the same time,the assignments of focal elements in these different groups corresponding to each source are added up respectively,in order to realize the mapping from the refined hyper-power set to the coarsened one.And then,two sources with the coarsened hyper-power set are combined together according to DSmC(Classical DSm combination rule) and PCR5(Proportional conflict redistribution No.5).The fused results from different groups will be saved as the connecting weights between father and children nodes.And then,all assignments of focal elements in different groups will be normalized respectively.Tree depth is set,in order to decide the iterative times in hierarchical system.Finally,by comparing new method with old one from different views,the superiority of new one over old one is testified well.