voidly

BelarusBY

falling (-10.4pp)

30-day day-by-day shutdown-risk trajectory. Each point is a per-horizon XGBoost prediction (h=1..30) with a 90% conformal band derived from absolute LOCO residuals. Peak risk highlighted.

Peak day
81.3%
d11 · 2026-06-19
Floor day
21.0%
d30 · 2026-07-08
30-day delta
-10.4pp
d30 vs d1 probability

30-day trajectory with 90% conformal band

model trajectory_v1
0%25%50%75%100%d1d7d14d21d30peak d11 81%
mean P(shutdown by day)90% conformal bandpeak day

Drivers (SHAP top 3 from the 7-day model — same features feed all horizons)

day 131.3%
2026-06-09
  • week_of_year-0.025
  • block_rate_roll14_mean-0.019
  • gdelt_unrest_30d+0.013
day 761.0%
2026-06-15
  • week_of_year-0.025
  • block_rate_roll14_mean-0.019
  • gdelt_unrest_30d+0.013
day 1442.6%
2026-06-22
  • week_of_year-0.025
  • block_rate_roll14_mean-0.019
  • gdelt_unrest_30d+0.013
day 3021.0%
2026-07-08
  • week_of_year-0.025
  • block_rate_roll14_mean-0.019
  • gdelt_unrest_30d+0.013

Honest caveats

  • Per-horizon XGBoost stacked into a trajectory — not a true joint distribution.
  • Conformal bands are absolute-residual q90 from LOCO predictions on 18 eval countries.
  • Long horizons (h>=14d) have wider effective uncertainty than the conformal q90 suggests; the same lag features feed all horizons.
  • Bands floor at [0,1]; probabilities below 0.05 are at the model's noise floor.
  • SHAP attribution above is from the 7-day point model. The trajectory uses 30 separate per-horizon XGBoost models that share the same feature set; we have not yet computed per-horizon SHAP.

Sample horizons

DayDateP(shutdown)90% interval
d12026-06-0931.3%[0%, 75%]
d72026-06-1561.0%[0%, 100%]
d142026-06-2242.6%[0%, 100%]
d302026-07-0821.0%[0%, 99%]
Spotlight countries