Selection and Assessment of Diabetes Risk Factors with Data Analy

សៀវភៅនេះនិយាយអំពីអ្វី?

This paper from The 14th National Conference on Information Technology 2022 is by Sodavy Tong and Nopadon Juneam of the Department of Computer Science, Kasetsart University, and Dangjai Souvannakitti of Phramongkutklao College of Medicine. It selects and assesses diabetes risk factors using physical examination records of 453 people from the Royal Dunyapabbumbad Foundation. The authors apply analysis of variance (ANOVA) to find factors linked to traditional diabetes risk scores, then compare three risk-factor sets of 17, 12 and 9 features with Random Forest, k-Nearest Neighbors, Logistic Regression, Support Vector Machine and Multilayer Perceptron models, judged by ROC curves and AUC. The set built from categorical features alone came within 10% of the traditional assessment. Statistics and data analysis students can read it as a worked example of feature selection and model evaluation. It can be read online and downloaded free of charge.