Research And Development Of Diabetes Prediagnosis System Based On Improved ABC And Apriori Algorithm

The great changes in people's lifestyles and the improvement of living standards have led to a younger age of onset of diabetes and a sharp increase in the prevalence.Diabetes-induced diabetic retinopathy,kidney and neuropathy,cardiovascular disease and other complications,also seriously threaten human health and life.On the other hand,with the continuous development of Internet technology and the continuous improvement of medical informatization,many hospitals have accumulated a large amount of medical data.Due to the large scale of data,unequal structure and diverse sources,traditional data analysis methods are difficult to extract valuable information efficiently.Therefore,based on the improved ABC and Apriori algorithm,this paper extracts and correlates the massive diabetes related data,designed and developed a diabetes auxiliary diagnosis system to assist the medical staff in diagnosing the diabetes,the system also provides an valid way for users to diagnose themselves.The main research of this paper were as follows:(1)The improved algorithm was used to select the characteristics of the diabetes feature set,as for deficiencies of observating bees' weak local search ability in the classic ABC algorithm,the observating bees' optimization mechanism was improved through local search strategy which can further refine and expand the search scope of the solution.The classical and the improved algorithm was compared by experiments.It was found that the improved algorithm can effectively filter the redundant features and has significant improvement in the quality,convergence speed and classification accuracy of the feature subset.(2)The Apriori algorithm was used to analysis the high-risk factors of diabetes,to avoid the classic algorithm's repeatedly scan the database which resulting in a lot of time spent on I/O operations,and frequent redundancy when the frequent itemsets connected itself,the transaction database was converted into a boolean matrix,and the calculation method is improved by compressing the matrix.The classical and the improved algorithm was compared by experiments.It was found that the improved algorithm runs faster and has less item set redundancy while under the same minimum support.(3)A diabetes-assisted diagnosis system based on the improved ABC and Apriori algorithm was designed and developed,and the unit and integration test of each module was carried out,finally the expected goal of auxiliary diagnosis was realized.It not only improves the diagnostic efficiency and accuracy of medical staff,but also provides timely and accurate disease warning for self-diagnosis patients.