Analysis And System Design Of Pathogenic Factors Of Hyperlipidemia Based On Feature Processing And K-C4.5 Algorithm

Hyperlipidemia often causes a variety of complications such as atherosclerosis,coronary heart disease,cerebral infarction and so on.It is a multiple chronic disease that seriously affects the health of residents in China.Therefore,early detection and prevention of this disease are extremely important.Aiming at improving the diagnosis efficiency and accuracy of hyperlipidemia among medical staff,this paper analyzed and studied the application of feature selection and feature extraction algorithms as well as classification algorithms in medical diagnosis,and designed and developed a hyperlipidemia auxiliary diagnosis system based on BPCA feature processing algorithm and K-C4.5 classification algorithm.The system can effectively assist medical personnel to diagnose hyperlipidemia.The main tasks as follows:(1)Aiming at the problems of existing feature processing algorithms in the processing of high-dimensional data,such as data loss,redundancy and poor noise removal,a BPCA feature processing algorithm based on improved BP neural network algorithm and principal component analysis algorithm is proposed.By redefining weights and thresholds,the network correction time and accuracy of the BP neural network when processing large amounts of data are improved,while reducing data redundancy and noise more efficiently;by improving the data reading process of the principal component analysis algorithm,repeated scanning of the database during data reading is avoided,and I/O operation time and data reading redundancy during data reading are reduced.Finally,by comparing the experimental results of BPCA algorithm,BP algorithm,and PCA algorithm through experiments,the BPCA algorithm model not only shortens the feature subset selection time,improves the accuracy rate,but also has high stability.(2)Aiming at the problems of existing feature processing algorithms such as data loss,redundancy and poor noise removal when processing high-dimensional data,a BPCA feature processing algorithm based on improved BP neural network algorithm and principal component analysis algorithm was proposed.By redefining the weight and threshold calculation formulas,the network correction time and accuracy of the BP neural network when processing massive data are improved,and the data redundancy and noise are more efficiently reduced;the data reading process of the principal component analysis algorithm is improved To avoid repeated scanning of the database when data is read in,reduce I/O operation time and data read redundancy when data is read in.Finally,through experimental comparison of the experimental results of BPCA algorithm,BP algorithm,and PCA algorithm,the BPCA algorithm model not only shortens the selection time of the feature subset,improves the accuracy rate,but also improves the stability of the algorithm.(3)According to the design needs of the hyperlipidemia auxiliary diagnosis system,a hyperlipidemia auxiliary diagnosis system based on the BPCA feature processing algorithm and the K-C4.5 classification algorithm is designed,and the function of each module of the system is developed and tested.The results show that the function of the system reached the expected goal,and the auxiliary diagnosis of hyperlipidemia is realized.