Top 1%
in this catalogue (246 scored practices)
Scores cluster high, so position within the catalogue is often more telling than the number alone.
Evidence of impact / measured public value3/3
Peer-reviewed in Environmental Health Perspectives with a documented prospective validation against the severe 2013 outbreak (forecast peak 863 vs observed 842 cases, 17-24% MAPE), independently summarised in a second PMC-indexed article.
Transparency, fairness & accountability3/3
The full forecasting methodology, predictor set, and validation results were published in a peer-reviewed academic journal, openly documenting both accuracy and error rates.
Transferability / demonstrated replication2/3
The LASSO-based method and its evaluation are published academically and have been benchmarked in follow-on studies (2024-2025), but the deployed system itself remains specific to Singapore's own surveillance data.
Scalability beyond pilot3/3
Deployed islandwide since 2013 and extended to roughly 50,000 Gravitrap devices with AI-based species/sex classification, indicating sustained national-scale operation.
Governance, capability & sustainability2/3
NEA operates the system in ongoing partnership with two universities and continues academic cross-validation, though no independently published audit of recent-year false-alarm or missed-outbreak rates was found.
Evidoria. Sistema de Alerta Precoce de Dengue com IA — Previsão de Surtos por Aprendizagem Automática da NEA de Singapura. Persistent ID: 9af44d22-d55f-4c78-8ca6-d06a38bac13d.