Machine-Learning-Based Diabetes Mellitus Risk Prediction Using Multi-Layer Neural Network No-Prop Algorithm

• DOI: 10.3390/diagnostics13040723
• Publication Date: 2023-02-15T08:37:24Z
• Authors: J. Jeba Sonia; Prassanna Jayacha...; Abdul Quadir Md; Senthilkumar Mohan; Arun Kumar Sivaraman; Kong Fah Tee
• Keywords: Confusion matrix

Over the past few decades, prevalence of chronic illnesses in humans associated with high blood sugar has dramatically increased. Such a disease is referred to medically as diabetes mellitus. Diabetes mellitus can be categorized into three types, namely types 1, 2, and 3. When beta cells do not secrete enough insulin, type 1 develops. create but body unable use it, 2 results. The last category called gestational or This happens during trimesters pregnancy women. Gestational diabetes, however, disappears automatically after childbirth may continue develop diabetes. To improve their treatment strategies facilitate healthcare, an automated information system diagnose required. In this context, paper presents novel classification using multi-layer neural network no-prop algorithm. algorithm uses two major phases system: training phase testing phase. each phase, relevant attributes are identified attribute-selection process, trained individually manner, starting normal then finally healthy Classification made more effective by architecture network. provide experimental analysis performances diagnoses terms sensitivity, specificity, accuracy, confusion matrix developed. maximum specificity sensitivity values 0.95 0.97 attained suggested With accuracy score 97% for categorization mellitus, proposed model outperforms other models, demonstrating that it workable efficient approach.