Loading investigation data...
ROC Curves
Receiver Operating Characteristic — higher area is better
Calibration Plot
Predicted probability vs. observed frequency
ROC Curves
Receiver Operating Characteristic — higher area is better
Calibration Plot
Predicted probability vs. observed frequency
Correlation Comparison
Pearson r with chronological age (forward prediction) — higher is better
Feature Importance (Hybrid Model)
How much each input contributes to the best-performing model
ROC Curves
Three models compared — domain, AI, and hybrid
Calibration Plot
Predicted probability vs. observed frequency
Feature Importance
What Matters Where?
Feature importance across six investigations. Click a column header to sort. Hover a cell for exact values.