WHO FluNet activity exceedance
Research task: predict whether a country's influenza test positivity will exceed 10% four weeks from now - a retrospective, observed-outcome forecasting task over WHO FluNet surveillance.
Versioned research data. Read the suitability requirements before interpreting scores.
Suitability and evaluation
Retained for retrospective model-development research with documented timing limitations. Later evaluation rows expose earlier outcomes; predictions require restricted access.
Supported uses: Retrospective train/validation model-development experiments under the documented timing and selection limitations; Final evaluation only under a protocol that isolates holdout data from development and prevents later rows influencing earlier predictions
Not supported: Claims of verified real-time or operational forecasting validity; Treating development/validation scores as untouched final-test results; Forecasting scores from unrestricted cross-row access to this artifact
- Historical feature availability
- assumed
- Training-label timing
- unverified
- Data vintage
- current snapshot
- Row selection
- unverified
- Required evaluation access
- chronological
Cross-row audit: recovery demonstrated. Independent operator-commissioned audit; predictions frozen before labels were opened; holdouts sealed; two AUC implementations in exact agreement
The labeled eval split is development/validation data. Isolate final holdout data from development and prevent later rows influencing earlier predictions. Published data is not a secure hidden test service.
- join (country, W) -> (country, W+28 days): the next row's inf_all_w and denom_w are the operands of T's label
- The labelled holdout is presumed to carry the same self-answering structure by construction; it was never opened by the audit and must be treated as compromisable by the same join class.
- Historical feature/label availability at the declared instant is assumed from current snapshots, not verified against release vintages.
- The current scorer passes the complete feature frame to predict_proba; prediction-time isolation is instructed but not mechanically enforced, so any framework score on these files is gameable.
- Splits use prediction week W although labels concern W+4; no label-availability embargo is enforced across train/validation or validation/holdout boundaries.
Corrections to historical documentation
- Public site page, Hub card and repository README must state the restricted suitability once rendered; no frozen package bytes change.
Reviewed 2026-10-24. Applies to this artifact; original data and tags are preserved.
At a glance
| Question | Research task: predict whether a country's influenza test positivity will exceed 10% four weeks from now - a retrospective, observed-outcome forecasting task over WHO FluNet surveillance. |
|---|---|
| Field | Infectious disease surveillance / binary-classification |
| Rows | 26,093 |
| Documented splits | 3 |
| This version | v2026.10 |
| Released | 2026-10-21 |
| Evaluation ROC AUC | 0.87 |
| Artifact | b1323a27fa02eb79de5fbb12c54068699ae15b3d44a9488942fd17ef0a4caffa |
Get this dataset
How to load it
from datasets import load_dataset
ds = load_dataset('earino/who-flunet-activity-exceedance', revision='v2026.10')
The version links and checksums identify this artifact. Exclude the labelled holdout from the agent's workspace when running an evaluation.
What is inside 
| Split | Rows | Positive labels |
|---|---|---|
| eval | 6,831 | 1,636 (23.95%) |
| holdout | 13,483 | 5,482 (40.66%) |
| train | 5,779 | 3,420 (59.18%) |
Files and checksums
| Download | Bytes | SHA-256 |
|---|---|---|
| meta.json | 2,915 | 592552716187b6645070fd625e1c197390a7f12fb1a719b9d8901ec0918f562b |
| private-holdout.csv | 3,482,298 | f7dd814a73c864df9d11ff2a1c39a899c97c9efe46b6687cb279c0d3c3ef276e |
| public-eval.csv | 1,589,403 | 7d809b16ed3b833bf13070b50a5f9fba83e3d2fb7569b993408e885eae9ad8e9 |
| public-train.csv | 1,496,272 | 2515289b43bdd49b51479b076f81a9f21bc2ec4382f3d0f7153687a8ecafb102 |
| quality.json | 5,610 | efa583ba6a07c1081488d9fe48d74736a484b9f48cd2d5712e2b0cbf190a062b |
Release notes · 2026-10-21
WHO publishes weekly influenza laboratory surveillance by country. The contribution here is not access: it is the frozen country-week panel, the declaration that the prediction instant is the end of week W, and the derived label - will positivity in week W+4 exceed 10%.
The release ships a reproducible baseline at eval AUC 0.8700. Two independent checks are published with it, because they decide what that number means. A calendar-only model reaches 0.6269, so the task is not merely seasonal. But week-W persistence alone reaches 0.8929 - a one-line rule beats the shipped baseline. That is stated on the card, not left to be discovered.
This is a retrospective, observed-outcome benchmark. It is not a prospective forecast, it is not calibrated, and the information cutoff rests on an assumption about publication timing that the publisher does not expose. The source's licence is unresolved: no open licence is asserted, and the terms are stated plainly.
Recorded structural checks and baselines
PASSED - gate 1.3.0, 32/32 checks, 0 failed, on the full artifact on the worker; staging release 418940777; hand-off --accept ACCEPTED (2026-10-20) Gate 1.3.0.
baseline only, through the runner's own contract. No agent or harness comparison was run, and none is implied.
These historical checks do not establish availability at each prediction time or protection against every shortcut. Scores below describe the recorded protocol.
Limits
Derived research dataset, not an official WHO product; WHO has not reviewed, authorised or endorsed it.
Licence unresolved: WHO states no dataset-specific licence for FluNet; terms for WHO data compilations require authorisation for alterations. No open licence is asserted. See LICENSE.md.
Retrospective observed-outcome forecasting. No claim of prospective operational validity.
The historical publication timing is unverified; the week-W information cutoff is an assumption, not a reconstruction of data vintages.
Row eligibility depends on whether the target week was later observed - information unavailable at prediction time.
Sentinel and non-sentinel surveillance are summed; 46% of source rows carry no origin label, so reporting-composition changes can look like activity changes.
The shipped baseline is below the persistence reference (0.8700 vs 0.8929); it is a reproducible baseline, not a best-known solution.
The holdout is labelled and published by design; keep its labels outside an evaluated agent's workspace.
No claim of real-time or operational forecasting validity: historical feature and label availability at the prediction instant is assumed from a current source snapshot, not verified against release vintages.
The labeled eval split is development/validation. Final holdout scoring requires prediction-time access restrictions; the original scorer provided the complete feature frame. Downloadable holdout labels must stay outside model development.
Source, licences, and citation
WHO FluNet - Global Influenza Programme (World Health Organization, Geneva)
World Health Organization, FluNet (Global Influenza Surveillance and Response System), https://www.who.int/tools/flunet, data retrieved 2026-10-20.
Source: NOASSERTION. Compilation: NOASSERTION. Code and documentation: MIT.
Our contribution - the row selection, the derived binary label, the temporal partition, the gating rules and the packaging - is our work, but it is not dedicated under any licence here. We lack the standing to apply CC BY-NC-SA 3.0 IGO (that licence may only be applied by an intergovernmental organisation) and we hold no written authorisation from WHO for the alteration. Nothing in this release grants you rights over the source rows, and nothing here may be presented as CC0, public domain, an open licence, or an official WHO product.
Suggested citation
World Health Organization. FluNet. Geneva: World Health Organization; 2026. Derived dataset: WHO FluNet influenza-positivity exceedance prediction task. This is a derived research dataset; it is not an official WHO product.