Measured evaluation results

Numbers below match this project's evaluation reports. Dataset scope, baselines, and limitations are documented under Integrity.

Test F1

0.800

Baseline 0.222

ROC-AUC

0.944

Baseline 0.901

PR-AUC

0.781

Baseline 0.323

Precision

0.952

Baseline 0.571

ROC Curve · AUC 0.944 (test split)

Precision–Recall · AUC 0.780 (test split)

Feature Importance · LightGBM (production model)

Curves and importances computed from the production LightGBM artifact on the chronological test split — same run as `reports/metrics/test_metrics.json`.

Metric	Primary	Baseline
Test F1	0.800	0.222
ROC-AUC	0.944	0.901
PR-AUC	0.781	0.323
Precision	0.952	0.571