THE ENGINE
CI-EWS Malaria Model
A LightGBM forecaster of confirmed malaria cases at district × month, trained on real CAR surveillance + climate indicators, with conformal prediction intervals — chosen by a head-to-head bake-off and evaluated on a 12-month hold-out.
00Deployed
Live forecast — LightGBM
Confirmed OPD malaria cases per district × month. Forecast for 2026-01 (data through 2025-12); seasonal-naive (climatological mean).
WIS
—
MAE
—
median abs. error
MASE
—
95% coverage
—
interval calibration
National forecast · 2026-01
216,036
cases · 95% 49,388–413,188
National reported cases — trailing 24 months
— reported cases
Per-district forecast · 2026-01
| District | Forecast | 95% interval |
|---|---|---|
| Carnot-Gadzi | 16,091 | 10,043–22,139 |
| Bambari | 12,772 | 1,524–24,020 |
| Alindao-Mingala | 11,391 | 0–23,532 |
| Bangui Iii | 10,692 | 2,650–18,734 |
| Nanga-Boguila | 10,296 | 1,599–18,993 |
| Paoua | 9,544 | 0–21,876 |
| Nana-Gribizi | 8,922 | 2,998–14,846 |
| Bimbo | 8,436 | 4,297–12,575 |
| Batangafo-Kabo | 8,133 | 0–21,176 |
| Bangui Ii | 7,531 | 0–17,097 |
| Sangha-Mbaere | 7,435 | 3,175–11,695 |
| Bossembele | 7,366 | 1,356–13,376 |
| Begoua | 6,727 | 1,943–11,511 |
| Kemo | 6,346 | 3,881–8,811 |
| Kembe-Satema | 6,294 | 0–15,022 |
| Bouar -Baoro | 6,083 | 3,056–9,110 |
| Mobaye-Zangba | 5,806 | 0–15,211 |
| Ouango-Gambo | 5,747 | 0–16,326 |
| Bangui I | 5,435 | 1,731–9,139 |
| Boda | 5,419 | 1,772–9,066 |
| Berberati | 5,416 | 0–12,295 |
| Bossangoa | 4,858 | 2,454–7,262 |
| Baboua-Abba | 4,777 | 951–8,603 |
| Haute-Kotto | 4,526 | 2,745–6,307 |
| Bangassou | 3,659 | 0–9,632 |
| Bozoum-Bossemptele | 3,559 | 0–8,485 |
| Kouango-Grimari | 3,155 | 0–7,105 |
| Mbaiki | 3,127 | 1,460–4,794 |
| Bocaranga-Koui | 3,015 | 0–6,767 |
| Haut-Mbomou | 3,001 | 1,079–4,923 |
| Bouca | 2,727 | 32–5,422 |
| Gamboula | 2,549 | 0–6,639 |
| Ngaoundaye | 2,369 | 0–5,191 |
| Bamingui-Bangoran | 2,357 | 642–4,072 |
| Vakaga | 475 | 0–1,436 |
01Registry
Model registry
MLflow registry unreachable right now — showing forecast + bake-off data only below.
04Design
Demand forecasting model
A planned model, not a shipped one — this documents what exists today, what's a placeholder, and what's still needed.
ILLUSTRATIVE — DESIGN SPEC
This model does not exist yet. Two of the six pieces below are real and already shipped (the case forecast and current stock feeds); the conversion step and output shape are running as an explicitly-tagged placeholder preview on the Supply Chain page today; the lead-time, safety-stock and uncertainty-propagation logic are not built at all.
Input
Case forecast (real, exists today)
The deployed LightGBM model already produces a per-district, next-month point forecast with a 95% conformal interval — this is the exact input the demand model will consume.
LIVEInput
Current stock (real, exists today)
RDT on-hand and ACT/AL courses dispensed per district, read live from HMIS — the exact input for a gap calculation.
LIVEConversion
Cases → commodity demand (placeholder today)
Supply Chain's preview uses one flat national ratio (RDTs/ACT-AL per case). The real model needs this calibrated per district and per commodity, since consumption per case varies with testing algorithm, referral patterns, and stockage practice.
PLACEHOLDERLogic
Lead time & safety stock (not built)
Procurement and distribution take weeks. The real model needs a reorder-point policy: order when projected stock will fall below a safety buffer before the next resupply cycle completes — not just a same-month gap.
NOT BUILTLogic
Uncertainty propagation (not built)
Today's preview uses only the case forecast's point estimate. The real model should propagate the forecast's 95% interval through to a demand range, so a procurement decision reflects how confident the case forecast actually is.
NOT BUILTOutput
District readiness + national summary (preview shipped)
National and per-district readiness gap, refreshed monthly alongside the case forecast — the output shape is already live on Supply Chain, just running on placeholder logic until the pieces above are built.
PLACEHOLDER05Benchmarks
Where it stands
Real hold-out comparisons plus the published CAR baselines worth comparing against. Green = cleared. Amber = open comparison.
Open
Seasonal-naive
Malaria
Same month last year — the floor every model must beat.
Open
MASE < 1
Malaria
Mean absolute scaled error below 1 beats seasonal-naive.
Open
95% PI coverage
Malaria
Prediction intervals should cover ~95% of actuals.
Open
Rubuga 2024 (DLNM)
Malaria
Published DLNM across all 35 districts — the CAR malaria baseline to compare against at district grain.
Open
Semakula 2020 (BYM)
Malaria
Bayesian spatial model — the reference for district-level evaluation.