We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
Logistic regression prediction models and key influencing factors analysis of diabetes based on algorithm design.
- Authors
Li, Zhijian; Pang, Sulin; Qu, Hongying; Lian, Wanmin
- Abstract
This article focuses on the key influencing factors and prediction accuracy of diabetes. Nine test indexes were mainly considered: low density lipoprotein, triglyceride, total cholesterol, white blood cell, temperature, blood pressure, heart rate, blood sugar, and age. By designing data experiment method and logistic regression prediction algorithm based on fivefold cross validation, it is used to analyze odds ratio, full subset screening regression, cross validation, root mean square error and confusion matrix on 96 original data samples with 76 diabetes patients and 20 non-diabetes patients. Based on statistical test and innovation, two modeling ideas based on the combination of clinical experience and statistical test are proposed. Logistic regression model I with 2 parameter variables and Logistic regression model II with 5 parameter variables are, respectively, established to predict and compare the accuracy of five groups of different cross validation test sets. The prediction accuracy of the former is 93.7895%, and that of the latter is 91.7895%. This study found that age and blood sugar are the key influencing factors of diabetes. However, total cholesterol, temperature and white blood cell have little effect on diabetes. The research method has high application value and can provide scientific solutions for medical institutions to predict, analyze and early diagnose diabetes.
- Subjects
LOGISTIC regression analysis; FACTOR analysis; REGRESSION analysis; STANDARD deviations; PREDICTION models; LEUCOCYTES; BLOOD sugar
- Publication
Neural Computing & Applications, 2023, Vol 35, Issue 36, p25249
- ISSN
0941-0643
- Publication type
Article
- DOI
10.1007/s00521-023-08447-7