基于不确定度的基本概率赋值概率转换方法

基本概率赋值概率转换是合成证据用于辅助决策的主要方法之一。针对现有转换方法中存在的缺少合理转换依据，灵活性和客观性不能兼顾的问题，提出一种基于不确定度加权的概率转换方法。该方法将概率转换视为一种决策过程，选择原BPA与最特异BPA间的Jousselme距离作为不确定性量度，通过“假设-校验”的方法联合求解最特异BPA与转换概率。实验算例表明，所提出方法的处理过程与人们的决策过程一致，概率转换结果合理有效。

Transformation of basic probability assignment to probability based on uncertainty degree

The transformation of basic probability assignment to probability is one of the main methods for combinational
evidence to decision-making assistant. However, the existing methods are lack of rational criterion, and can not get flexibility
and objectivity simultaneously. Therefore, a transformation is presented based on uncertainty degree weighting. The method
takes the process as a decision-making problem, chooses the Jousselme distance between the original BPA and the most
special BPA as the uncertainty measure. The most special BPA and transformation probability are got through united solving
method called “assumption-verification”. The experimental examples show that the proposed method is agreed with our
decision-making custom, and the transformation result is reasonable and effective.