逐步回归与判别分析的应用研究——在乳腺疾病建模中的应用

Stepwise Regression and Discriminant Analysis in Breast Disease Modeling

体内激素紊乱是导致乳腺疾病发生的重要原因.将逐步回归法与Fisher判别法结合应用于乳腺疾病建模中,对乳腺疾病与6种内分泌激素及患者年龄的关系进行了多元线性判定分析,建立了变量与相关疾病的逻辑回归模型与判别模型.该模型通过实际应用,判别效果较好,有助于乳腺疾病的诊断.

Hormone disorder is an important cause leading to breast disease. This article combined Stepwise Regression and Fisher Discriminant in building a model of breast disease. The relationship between six kinds of endocrine hormone, patient＇s age and breast disease is analyzed through Multiple Linear Discriminant Method. In the end, for element variable and correlative disease, this article gets a logistic regression model and a discriminant model. Through application, the model is testified to be effective for the diagnosis of breast disease.