Multi-Class Classification Approach for Predicting Type 2 Diabetes Mellitus Prediction Using Integrated Deep Learning Network with Feature Engineering

Key Points

• Type 2 diabetes prediction accuracy improves significantly with deep learning techniques and feature engineering.
• The model achieved 85% predictive accuracy on the test dataset, indicating its potential for clinical application.
• Utilizing an integrated deep learning network, researchers assessed feature importance and optimized data processing.
• The findings underscore the ability of AI-driven models to facilitate early diabetes intervention, although real-world implementation is crucial.