EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Machine learning reveals temperature as a key predictor of dengue risk across Thailand’s provinces: A 20-year analysis

Machine learning zeigt Temperatur als wichtigen Prädiktor des Dengue-Risikos in Thailands Provinzen: eine 20-jährige Analyse

Pikkanet Suttirat, Sudarat Chadsuthi, Supassorn Aekthong, Joacim Rocklöv, Dominique J Bicout, Peter Haddawy, Myat Su Yin, Saranath Lawpoolsri, Charin Modchang iD

PLOS Neglected Tropical Diseases · 28.07.2026 · 20 (7) · e0014590

DOI: 10.1371/journal.pntd.0014590 ↗

Auf einen Blick

Ein Machine-Learning-Modell aus 20 Jahren Dengue-Daten aus Thailand fand Temperatur als wichtigsten Prädiktor. Die Vorhersageleistung war vor der Pandemie besser als während und nach der Pandemie.

Fragestellung

Welche Umwelt-, Klima- und sozioökonomischen Faktoren sagen das Dengue-Risiko in Thailands Provinzen voraus?

Methode

Machine-Learning-Analyse mit 20 Jahren monatlicher Surveillance-Daten (2003–2022) aus 77 Provinzen; 54 Merkmale; Vergleich von 4 Klassifikatoren; XGBoost gewählt; SHAP-Analyse.

Ergebnisse

XGBoost: AUC 0,80 im Pre-Pandemie-Test und 0,74 in der kombinierten During-/Post-Pandemie-Phase. Temperatur stellte 7 der 10 wichtigsten Merkmale; AUC 0,62 in 2021.

Einschränkungen

Leistung verschlechterte sich während der COVID-19-Pandemie; systematische Übervorhersage zeigte Einflüsse außerhalb des Modells.

Schlussfolgerung der Autor:innen

Temperatur war der dominierende Prädiktor für Dengue-Risiko in Thailand. Umweltbasierte Vorhersagen sind unter stabilen Bedingungen zuverlässig, nehmen bei gesellschaftlichen Störungen aber deutlich ab.

Kernaussagen

  • Temperatur dominierte die Merkmalsrangfolge, 7 der 10 wichtigsten Merkmale.
  • AUC 0,80 vor der Pandemie, 0,74 in der kombinierten During-/Post-Pandemie-Phase.
  • Precipitation trug wenig bei; höheres Gross Provincial Product war mit höherem Risiko assoziiert.

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

Kurz gefasst

Fachartikel in der Zeitschrift PLOS Neglected Tropical Diseases, online erschienen am 28.07.2026. Verfasst von Pikkanet Suttirat und 8 weiteren Autor:innen. Beteiligte Einrichtung: Medical Biophysics & Disease Modeling Group, Department of Physics, Faculty of Science, Mahidol University, Bangkok, Thailand. Hauptthemen laut MEDLINE: Temperature, Dengue, Machine Learning. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 72 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Rockefeller Foundation, Mahidol University.

Auf einen BlickAbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Dengue fever remains a critical public health challenge in Thailand, with transmission dynamics driven by complex interactions between environmental and socioeconomic factors. Understanding these predictive factors is essential for developing robust forecasting systems.

Methods

We developed a machine learning framework to classify spatiotemporal dengue risk and identify key predictive factors across Thailand. We analyzed 20 years of monthly dengue hemorrhagic fever surveillance data (2003–2022) from 77 provinces, integrating 54 environmental, climatic, and socioeconomic features. We benchmarked four candidate classifiers — logistic regression, support vector machines, random forests, and eXtreme Gradient Boosting (XGBoost) — and selected XGBoost on the basis of performance across six metrics. SHapley Additive exPlanations (SHAP) were used to interpret feature contributions. The dataset was stratified into training (2003–2016) and testing periods, with the latter subdivided into pre-COVID-19 (2017–2019), COVID-19 (2020–2021), and post-COVID-19 (2022) phases.

Results

The XGBoost model achieved an AUC of 0.80 in pre-pandemic testing and 0.74 across the combined during- and post-pandemic period. Temperature dominated the feature-importance ranking, comprising seven of the top ten features, with non-linear thresholds near 21°C for 1-month lagged minimum temperature and near 32°C for 3-month lagged maximum temperature — values that align with established biological constraints on Aedes aegypti –mediated transmission. Precipitation features contributed minimally to model predictions, while a higher Gross Provincial Product was associated with increased dengue risk, consistent with predominantly urban transmission patterns. Model performance deteriorated significantly during the COVID-19 pandemic (AUC = 0.62 in 2021), with systematic overprediction indicating that non-environmental factors operating outside the model dominated dengue dynamics during this period.

Conclusions

Temperature is the dominant predictor of dengue risk in Thailand, and the thresholds we recover correspond closely to known biological constraints on vector competence. Environmentally driven prediction is reliable under stationary conditions but degrades substantially during periods of major societal disruption, underscoring the need to integrate behavioral and surveillance-coverage indicators alongside environmental predictors when applying such models in real time.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

TemperatureDengueMachine LearningThailandHumansBoosting Machine Learning AlgorithmsPredictive Learning ModelsRandom ForestAnimalsLogistic ModelsClassification AlgorithmsPrediction AlgorithmsAedesDengue Virus

Autor:innen und Institutionen

  • Medical Biophysics & Disease Modeling Group, Department of Physics, Faculty of Science, Mahidol University, Bangkok, Thailand.
  • Center for Disease Modeling, Faculty of Science, Mahidol University, Bangkok, Thailand.
  • Department of Physics, Faculty of Science, Naresuan University, Phitsanulok, Thailand.
  • Department of Public Health and Clinical Medicine, Section of Sustainable Health, Umeå University, Umeå, Sweden.
  • Heidelberg Institute of Global Health, Heidelberg University, Heidelberg, Germany.
  • Interdisciplinary Center of Scientific Computing, Heidelberg University, Heidelberg, Germany.

ORCID-Profile: Charin Modchang

Förderung

  • Rockefeller Foundation
  • Mahidol University

Zitieren

APA

Suttirat, P., Chadsuthi, S., Aekthong, S., Rocklöv, J., Bicout, D. J., Haddawy, P., Yin, M. S., Lawpoolsri, S., & Modchang, C. (2026). Machine learning reveals temperature as a key predictor of dengue risk across Thailand’s provinces: A 20-year analysis. PLOS Neglected Tropical Diseases, 20(7), e0014590. https://doi.org/10.1371/journal.pntd.0014590

Vancouver

Suttirat P, Chadsuthi S, Aekthong S, Rocklöv J, Bicout DJ, Haddawy P, et al. Machine learning reveals temperature as a key predictor of dengue risk across Thailand’s provinces: A 20-year analysis. PLOS Neglected Tropical Diseases. 2026;20(7):e0014590. doi:10.1371/journal.pntd.0014590

BibTeX
@article{Suttirat2026,
  author = {Suttirat, Pikkanet and Chadsuthi, Sudarat and Aekthong, Supassorn and Rocklöv, Joacim and Bicout, Dominique J and Haddawy, Peter and Yin, Myat Su and Lawpoolsri, Saranath and Modchang, Charin},
  title = {Machine learning reveals temperature as a key predictor of dengue risk across Thailand’s provinces: A 20-year analysis},
  journal = {PLOS Neglected Tropical Diseases},
  year = {2026},
  volume = {20},
  number = {7},
  pages = {e0014590},
  doi = {10.1371/journal.pntd.0014590}
}
Kostenlos starten →