Leveraging AHP and transfer learning in machine learning for improved prediction of infectious disease outbreaks
2024

Improving Predictions of Infectious Disease Outbreaks Using Machine Learning

Sample size: 1716 publication Evidence: high

Author Information

Author(s): Abdallah Reham, Abdelgaber Sayed, Sayed Hanan Ali

Primary Institution: Helwan University, Cairo, Egypt

Hypothesis

Can advanced machine learning techniques improve the prediction of infectious disease outbreaks?

Conclusion

The study demonstrates that an ensemble model achieves the highest accuracy for predicting Zika outbreaks at 96.80% and effectively balances precision and recall for Chikungunya.

Supporting Evidence

  • The ensemble model achieved the highest accuracy rate of 96.80% for predicting Zika outbreaks.
  • The model exhibited consistent performance across various metrics.
  • In the context of Chikungunya, the ensemble model achieved an accuracy of 93.31%, with a precision of 57% and a recall of 63%.
  • The study utilized a comprehensive dataset comprising climate and socioeconomic data from 2007 to 2017.

Takeaway

This study shows how computers can help predict when diseases like Zika and Chikungunya might spread, which can help keep people safe.

Methodology

The study used machine learning algorithms, including Random Forest, XGBoost, Gradient Boosting, and an ensemble model, along with the Analytic Hierarchy Process for feature selection.

Potential Biases

The study's reliance on historical data may introduce biases related to data quality and availability.

Limitations

The study faced limitations related to data availability, particularly for Chikungunya and Zika, and lacked comprehensive clinical data at the patient level.

Statistical Information

P-Value

p<0.05

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1038/s41598-024-81367-1

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication