Conference: ICAIII 2026
Publisher: IEEE

This research focuses on risk stratification and early complication prediction, moving beyond single-outcome prediction toward a system-level preventive approach. The framework combines clustering and supervised learning to support targeted intervention.

Key Contributions:

• Stratified diabetes patients into meaningful risk groups using clustering techniques
• Predicted early complications such as cardiovascular and renal risk
• Addressed class imbalance using robust resampling strategies
• Enabled prioritised, preventive healthcare planning
• The framework supports early intervention, improved resource allocation, and personalised care pathways.