Classification Variables Of Logistic Regression Model And Its Application Research

The article selects the quarterly financial datas of the gem listed companies from2010to2012,and choose to reflect the company’s profitability index,which have25indi-cators,including2838datas.In the process of the datas integration,the handle of missing data adopt the method,which use missing value that near the mean of it to instead the missing value.We base on analysis for logistic regression model of the gem listed companies profit,and discuss that use the single of logistic regression model,which is easy to appear the problems of Multicollinearity and model predict inaccurate,so the model is needed to improve.We establish comprehensive logistic regression model.By means of factors analysis and extract the6principal component analysis. Us ing logistic regression model establish the gem listed company’s profit model.After the inspection,common factor can explain the total factor94.26%,and make the overall prediction accuracy reaches to99.4%,which ex-plan that the logistic regression model is based on factor analysis and principal component analysis can react the enterprise profit objectively.