Research And Application Of Evidence Combination Method Based On Information Degree

Evidence theory is a commonly used uncertainty reasoning method,which can easily express and deal with the uncertainty of information with its complete mathematical basis,and evidence theory has its own advantages.However,when the classical evidence combination method is used,there will be a situation contrary to people's cognition,so that its application will be limited to a certain extent.Therefore,scholars at home and abroad have put forward a lot of improvement ideas from all angles.The method of evidence combination is further developed and optimized.On the basis of previous scholars' research,this paper focuses on the information between the sources of evidence,and proposes a method of evidence combination based on the degree of information.At the same time,the validity of the synthesis results is established to evaluate the information entropy of the synthesis results to reflect the effectiveness of the synthesis results.In the application of evidence theory,combined with the problem of classification.Based on the idea of evidence theory,a decision tree construction method based on evidence theory is proposed,and an example is given to verify the effectiveness of the proposed method.The evidence combination method based on information degree first considers the relationship between evidence,constructs the information degree to represent the relationship between evidence,modifies the evidence source by using information degree,and then synthesizes the evidence.The information entropy is used to reflect the effectiveness of the synthesis results.By comparing with the existing synthesis methods,it can be seen that the focusing property of this method is greatly improved,and the rationality of the synthesis results is guaranteed at the same time.On the other hand,in the application of evidence theory,when classifying continuous data,we fully consider the uncertainty of data,introduce evidence theory to construct the function model of transformation training set,and make use of the evidence synthesis method proposed in this paper.The transformed training set is synthesized by evidence,combined with the idea of decision tree construction,the definition of class attribute entropy and attribute entropy are established,and the decision tree based onevidence theory is generated,and the method is verified by data set.Compared with the classical decision tree algorithm,the effectiveness of this method is illustrated.