Research on Weighted Fusion for Prior Information Based on Entropy
Zhang Ming-qin · Microcomputer Information · 2009
The traditional prior information fusion methods have two disadvantages——subjectivity and complexity, which makes the result of the previous methods is unreliability for the reliability analysis with small sample. To resolve this problem, a method of prior information weighted fusion based on entropy is proposed, and the corresponded algorithm is designed. Then, the idea of prior information fusion verification based on posterior distribution is presented; Finally, according to the verification idea, the simulation experiment of the method is designed and realized using Matlab. The result shows the new method is more simple and straight for the prior information fusion and more objective for achieving the prior distribution than the traditional methods.