All five figures are downloaded directly from the pipeline repo at
results/figures/ (commit cf180acf). Click to open full-size PNG.
Observed vs Predicted (TEST window)
The full-feature negative-binomial GLM point predictions against actual weekly counts for the held-out 2022-W01 → 2023-W52 window.
Feature Importance
Top-20 features by absolute coefficient from the full-feature NB GLM, ordered for stability across re-fits.
Residuals (TEST)
Time series of (y - y_hat) on the test window, with shaded ±1σ bands from
the rolling-origin backtest residuals.
ROC Curve (binary ≥1)
ROC curves overlaid for the three models on the binary ≥1 task, evaluated
on the held-out test window.
Calibration (binary ≥1)
Reliability diagram — predicted probability vs observed frequency across 10 probability bins.
The figure files ship in results/figures/ once the pipeline runs make figures
on the next reproduce. The site rebuild pulls them straight from
results/figures/ in the project repo.