Research On Early Warning Of Type 2 Diabetes Based On Fuzzy Decision Tree

With the rapid development of information technology,the extensive accumulation of various data,complex decision-making system,data mining has become a business and scientific research and other needs,become a new research and application focus.In the medical and health field,medical data has increased dramatically due to the application of various medical information systems,the digitization of medical devices and the application of portable physiological testing equipment.These medical data have profound implications for disease management,control and medical research significance.The paper combines machine learning with diabetics data,specializes in data processing,provides personalized medical advice,forms an early warning mechanism for possible health problems,avoids high-risk patients becoming diabetics or deteriorates,and reduces the burden on individuals and health care,In order to achieve the scientific management of chronic diseases such as diabetes.Fuzzy sets and rough sets can deal with the information uncertainty in the knowledge expression system.In this paper,we introduce the theory of the positive and upper approximation set in the rough set into the concept of membership degree of fuzzy theory,and combine the two to form Fuzzy membership and fuzzy approximation sets,the knowledge of fuzzy set approximation set theory is used to solve the optimal fuzzy decision tree splitting attribute,and it is found that this method can deal with the uncertain information more effectively.The fuzzy decision tree algorithm based on fuzzy rough set proposed in this paper is compared with fuzzy ID3.The experimental data is based on the UCI data set.It can be seen from the experimental results that the fuzzy decision tree has a higher number of layers The number of trees constructed is smaller than that of fuzzy ID3,and the structure is more concise,which makes it easier for people to understand and make the right decisions quickly.Based on the analysis of diabetes mellitus diagnostic criteria,the roughness technique is used to reduce the attribute of diabetes mellitus.Then,based on the attribute reduction,the reduced attribute is fuzzified,and finally the fuzzy roughness Pretreatment of data after the construction of fuzzy decision tree,extract the rules,the formation of diabetes early warning mechanism.Diabetes can be based on the formation of health warning,timely adjustment of their eating habits,living habits,with a view to effectively reduce the incidence of diabetes.