Predicting Dengue Incidence Using Search Data
Author Information
Author(s): Benjamin M. Althouse, Yih Yng Ng, Derek A. T. Cummings
Primary Institution: Johns Hopkins Bloomberg School of Public Health
Hypothesis
Can internet search data effectively predict dengue incidence?
Conclusion
Internet search terms predict dengue incidence with high accuracy and may be useful in areas with underdeveloped surveillance systems.
Supporting Evidence
- The linear model selected was superior to other models considered.
- In Singapore, the model showed a correlation of 0.931 between predicted and observed dengue incidence.
- In Bangkok, the model showed a correlation of 0.869.
- SVM models outperformed logistic regression in predicting periods of high incidence.
Takeaway
This study shows that looking at what people search for online can help predict how many dengue cases there will be, which is important for keeping people safe.
Methodology
The study used internet search data and dengue incidence data from Singapore and Bangkok to create predictive models using various regression techniques.
Potential Biases
Search behavior may be influenced by media reports, which could affect the accuracy of predictions.
Limitations
The model's performance may vary in different settings, and it relies on the availability of internet search data.
Participant Demographics
Data were collected from Singapore and Bangkok, but specific demographic details of participants are not provided.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website