Cancer Risk Prediction โ€” AI Screening Tool

๐Ÿงฌ
Cancer Risk Prediction
AI-Powered Screening Tool ย ยทย  9 Machine Learning Models
โš•๏ธThis tool is for screening purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
Model Guide โ€” Which Should I Use?
โญ HighRecall_Ensemble โ€” Recommended
Best for screening. Combines 4 models with boosted minority-class weights to catch the most cancer cases. Lowest false-negative rate.
๐Ÿฅˆ Ensemble_L2_ET โ€” Strong Runner-Up
Logistic + ExtraTrees soft-vote. Better precision than the top model โ€” fewer false alarms.
๐Ÿ“ SVM โ€” Highest AUC (0.931)
Best at ranking patients by risk score. Use for risk stratification rather than binary decisions.
๐Ÿ“Š Logistic Models โ€” Most Explainable
L1 (feature selection) and L2 (stable coefficients). Best when transparency and interpretability matter.
Prediction Model
HighRecall_Ensemble is recommended for most users.
Biological Sex
1 10
First-degree relative (parent, sibling, child) diagnosed with any cancer.
e.g. Crohn's disease, rheumatoid arthritis, chronic hepatitis, IBD.
1 10