We predict dengue outbreak using weather data. Dengue data is collected from Directorate General of Health Services. Weather data is collected from Darksky API. To determine the degree of dengue outbreak (namely- high, medium, and low) in a region, we use a Convolutional Neural Network (CNN) based on the SuperTML approach. Alongside this method, we also apply the traditional classifiers. We find that the CNN using SuperTML achieves excellent performance with an accuracy of 0.74, which is significantly higher than the results achieved by the traditional classifiers.
Citation Request:
inproceedings{leon2022dengue, title={Dengue Outbreak Prediction from Weather Aware Data}, author={Leon, Mazharul Islam and Iqbal, Md Ifraham and Meem, Sadaf and Alahi, Furkan and Ahmed, Morshed and Shatabda, Swakkhar and Mukta, Md Saddam Hossain}, booktitle={Bangabandhu and Digital Bangladesh: First International Conference, ICBBDB 2021, Dhaka, Bangladesh, December 30, 2021, Revised Selected Papers}, pages={1--11}, year={2022}, organization={Springer} }
Leon, Mazharul Islam, et al. "Dengue Outbreak Prediction from Weather Aware Data." Bangabandhu and Digital Bangladesh: First International Conference, ICBBDB 2021, Dhaka, Bangladesh, December 30, 2021, Revised Selected Papers. Cham: Springer International Publishing, 2022.