Study On Hypertension Data Based On Principal Component Regression And Decision Tree

This paper selects sample data from the survey of hypertension epidemics among residents in rural areas of a certain city.The survey data includes three parts: baseline survey from 2004 to 2006,and two visits in 2008 and 2010,with a total of 46,315 records.This article selects the baseline survey of 2004-2006 to study the relation-ship between hypertension related factors and hypertension.There are many factors influencing high blood pressure,such as age,gend er,and eating habits.Because there are too many influencing factors,in order to facilitate the processing,this paper adopts the principle of principal compon ent analysis,and turns multiple indicators into a few unrelated comprehensive indicators to achieve the dimension reduction of the influencing factors.Next,regression analysis and logistic regression analysis were performed on the principal components to obtain a judgment model for hypertension.Also,perform data simulation.In the model obtained by principal component regression and principal component logistic regression analysis,the principal component logistic regression can be used to derive the probability of disease and make a good prediction.The above is based on the traditional multivariate statistical analysis of the processing of this set of data.At the same time,we also process the data based on the decision tree algorithm in modern popular data mining to obtain a discr-iminant model of hypertension.According to the model,it is possible to judge whether the residents have high blood pressure according to some living in dicators of the residents,and at the same time,according to this model,some suggestions for prevention and treatment can be given.This paper uses two models to find out some factors that need attention in life,and gives some effective suggestions for the prevention and treatment of hypertension in residents.And to achieve the application of mathematical models in practice.