Research On Optimizes Constructive Three-way Decisions Model And Application In EEG Signal

Three way decision theory is the expansion of the traditional binary decision model,it has three decision rules,such as positive,negative and boundary regions.Since the three way decision model proposed,it has been applied to analysis the uncertain or incomplete information.In medical diagnosis,the three decision model also had many applications.The three way decision model is proposed in the study of rough sets and decision rough sets,Initially,it was provide interpretation for the rough sets' three domains.The three way decision model based on DTRS could give a good decision results,but there has subjective factors in select of the loss function,so this model could not reflect the objective information of the decision data.Zhang proposed the three way decision model based on constructive covering algorithm(CCA),which solves the problem of obtaining the threshold values.Although,the CCA model solved the problem of obtain threshold,however,the covering center of construct the covering domains is randomly select.Considering the classified data generally have some regularity,that is,the data of the same class are always similar in spatial distribution.The random method of select covering center can not be form the optimal cover domains.Therefore,in this paper,firstly,we optimize the select covering center method of three way decision model based on CCA.Secondly,we applied the optimized three way decision model(O_CCA)in EEG signal analysis of Parkinson's disease.The optimize covering center select method mainly by introducing the nearest mean theory.It is highlight of this paper to apply the O_CCA model to the analysis the EEG signal.The traditional binary classification models don't accord with the principle of diagnosis of doctors.However,the three way decision model based on O CCA combines the discrete wavelet transform and sample entropy as analysis scheme.This scheme can make exact analysis in clinical diagnosis,and provides important basis for the doctors.This paper mainly studies the theory and application of three way decision model,and the main work includes:1.This paper firstly introduces the proposed,development,and application of the three way decision model,and elaborate the three way decision theory in detail.Then,reviewed the knowledge and research on analysis of EEG signals of Parkinson's disease,and studies the analysis of EEG signal detailedly.At last,combine the nearest mean theory and three way decision model based on CCA,we proposed the optimize covering center three way decision model.2.The three way decision model based on O_CCA solved the problem of randomly selecting the covering center,enhanced the accuracy of classification and the classifier effect is better.In EEG analysis,the binary classifier had been pervasive application.And when the data is limited,the binary classifier model can't provide much more information for the patient to diagnosis.The condition of delay decision actually exist According to the three decision strategy such as acceptance,rejection,and non-commitment,we study the three way decision model applied in diagnosis.3.The three way decision theory's domain corresponding to the diagnosis completely.In this paper,we applied the three way decision model based on O_CCA in EEG signal analysis to distinguish the subject has Parkinson's disease or non or the information is insufficient.We proposed a scheme combine the DWT,SampEn,and three way decision model based on O_CCA in analysis EEG.The experiment shows that the scheme we proposed has better result.