Research On Diagnosis And Treatment Of Tcm Based On Adaptive Fuzzy Neural Network For Sleep Disorder

Traditional Chinese medicine is an important achievement in the medical field that the Chinese nation has passed down for thousands of years.It passes the method of "seeking,hearing,inquiring and cutting" four diagnostics,dialectical treatment of human body.Therefore,the traditional Chinese medicine diagnosis method has a large degree of subjectivity and fuzziness and is not easy to master.This traditional Chinese medicine diagnosis method has hindered the development of Chinese medicine,It is necessary to use the technology of artificial intelligence to make dialectics "objective." The following two intelligent diagnostic methods are commonly used:Artificial neural network diagnosis method: ANN has strong self-learning and association functions.It is good to use expert knowledge,but it can't handle and describe fuzzy information,the existing knowledge and experience cannot be used to explain the reasoning process and reasoning results of disease diagnosis.Based on expert system: The fuzzy diagnosis reasoning method can not only deal with the determination of symptoms,but also can deal with uncertain fuzzy symptoms,improve the ability of the symptom diagnosis system to process language knowledge.The disadvantage is that they cannot acquire knowledge by themselves,poor self-learning ability,unable to automatically generate rules and adjust membership functions,the knowledge base is difficult to maintain.This article uses ANFIS to study the diagnosis and treatment of sleep disorders in Chinese medicine.ANFIS(Adaptive Fuzzy Neural Network)integrates learning,association,recognition,and information processing,drawing on the advantages of the above two diagnostic methods,and make up for their shortcomings.At present,adaptive fuzzy neural networks have not seen relevant literature in the diagnosis of sleep disorders in Chinese medicine.This article uses ANFIS to study the diagnosis and treatment of sleep disorders in Chinese medicine.The choice of disease symptoms is important for diagnosing a disease,choosing a range of symptoms that is too large can make it difficult to choose a case,choosing too small will miss out on symptoms that are important for diagnosis.Therefore,this paper proposes mutual information based on mutual information(It refers to the correlation between two sets of variables,if the two variables are independent of each other,mutual information is zero;If the two variables depend on each other,the mutual information is the largest.)genetic algorithm tocompress clinical symptoms,from the point of view of Shannon's information theory,Using the symptom-symptom mutual information to characterize the symptomatic information content in the symptoms,symptom-symptom information is used to characterize the degree of independence between syndromes.By constructing a comprehensive mutual information indicator,using the genetic algorithm to select a subset of the symptoms of the first symptom,the symptom,the high mutual information,and the small symptom-information information.achieve compression of clinical symptoms.In order to be able to perform simulation calculations,it is also necessary to normalize the main symptoms of compression.According to Mamdani and T-S fuzzy reasoning,traditional Chinese medicine theory and doctor's clinical experience,constructing a syndrome-based diagnosis model of sleep disorders in traditional Chinese medicine based on adaptive fuzzy neural network,the data normalized to the four types of symptoms is used as the input value of the diagnostic model,The output of the model is the four types of syndrome results corresponding to the sleep disorder of Chinese medicine.research shows: data trained by the diagnostic model,the result of the output afterthe adjustment is displayed,the difference from the original value is small.Therefore,ANFIS is feasible to diagnose TCM sleep disorders.And more powerful than other diagnostic models.