Research And Development For Online Classification Of Hypertension System Based On The Binomial Logistic Regrseeion

The improvement of material life and the change in eating habits,which caused some non communicable disease that threat increasingly the health of human beings.In the paper,we will distill some data about hypertensive patients online,mining and analyzing some of the hidden information,processing and classifying to get a comparatively accurate classification results.The result will make patients have more precise knowledge of their disease and achieve the best therapeutic way.The aim of this paper is the construction of classification model which be based on condition description of hypertensive patients online and the treatment plan from doctor.Then we will develop a online diagnosis system for hypertension by the model.The system can be divided into three types of hypertension in the following categories:0(no disease),1(minor),2(severe).The research process of this paper consists of data preprocessing,model building and system development.The details are as follows:(1)Classifying the words of textual data;(2)Distilling some attributes from the description of patients(eg age,sex,etc),at the same classify the treatment plan from doctor;(3)Converting data and translating textual data into numerical data;(4)Building two models which the first be used to divide the condition of patients into no disease or having disease and the second be used to divided the prevalence into minor.and severe;(5)Developing a system based on the first and second model.The developed system based on binary logistic regression model not only have a precise forecasting about the prevalence of patients,but also have some theoretical and practical significance on medical service system online.