摘要
通过对年龄、BMI、家族史等19个因素,三种不同性质变量研究,利用相关性分析,得到糖尿病家族史、脂肪、文化程度和收缩压等相关程度较大的四个因素,并建立患病概率预测模型,该预测结果在疾病的预防,监视方面以及医疗保险中保单的核保方面有一定的指导意义。
The paper focuses on three different types of variables and nineteen factors leading to diabetes including age,BMI,family heredity, etc.Based on correlation analysis,we regard that the most relevant factors are family heredity,fat,education and systolic blood pressure. Meanwhile,the sick probability model is imitated.The result of the paper has a strong operation and guiding significance in the precaution and surveillance of diseases and medical insurance.
出处
《湖州职业技术学院学报》
2014年第1期76-78,83,共4页
Journal of Huzhou Vocational and Technological College
关键词
糖尿病
相关性分析
概率神经网络
diabetes
correlation analysis
probability neural network
