Research And Design Of Disease-aided Diagnosis Software Based On Improved RF-LR Algorithm

In recent years,with the introduction of the "Healthy China" policy,medical and health issues have received widespread attention in the society.Due to the current medical resource distribution is unbalanced and the disparity in the level of civic and country medical services,research and application in the field of disease-assisted diagnosis has important practical significance.At present,there are problems in the field of diseaseassisted diagnosis that the accuracy of the disease prediction algorithm is not high enough,the prediction algorithm does not take into account the cost of disease misdiagnosis and the backwardness of traditional disease-assisted diagnosis software development tools.In view of the above research status,this study presents a disease prediction model based on the random forest and logistic regression(RF-LR)improved algorithm,uses real medical data to test and analyze the algorithm,and designs and implements disease-assisted diagnosis software based on the proposed algorithm.The main work of this article is as follows:1.Build a disease predictive model based on random forest and logistic regression(RFLR)improved algorithm.To solve the problem of the accuracy of disease prediction algorithms is not high enough,a feature selection method based on random forest and sequence backward search is used to eliminate redundant features in the data set and improve the accuracy of the algorithm.In view of the cost of disease misdiagnosis,a costsensitive learning method is used in the logistic regression algorithm.The cost weight parameter is added to the loss function of logistic regression.By selecting the optimal weight parameter,the cost of disease misdiagnosis is reduced.Using real medical data,the improved algorithm is compared with logistic regression,decision tree,and support vector machine.The test results prove that,compared with other algorithms,the performance of the disease prediction algorithm proposed in this study is better,and the accuracy rate,recall rate and F1 value are 89.3%,86.8%,and 88%.2.The hierarchical design of the disease-assisted diagnosis software divides the software into five layers: presentation layer,control layer,business logic layer,data persistence layer and data layer.And using the B / S architecture mode to build the software.According to the analysis of software needs,Struts,Spring,Hibernate(SSH)framework is used for software design and build.The functions of the disease-assisted diagnosis software include user registration and login,user management,disease prediction,disease consultation,and popular science push.At the same time,the improved prediction model is used in the disease prediction function.3.According to the user’s demand for disease-assisted diagnosis and the national software quality testing standards,a software testing environment is built to test software,including software function testing software compatibility testing and software performance testing.According to the test results,the software can satisfy the user’s demands of disease-assisted diagnosis.At the same time,the software complies with national software standards,and can be run and used steadily in different browsers.