Utilização de técnicas de Machine Learning e de Deep Learning para a predição de casos de internações causadas por dengue em municípios da Paraíba

• Ewerthon Dyego de Araújo Batista UEPB / IFPB
• Wellington Candeia de Araújo UEPB
• Romeryto Vieira Lira IFPB
• Laryssa Izabel de Araújo Batista UFPB

DOI: https://doi.org/10.5753/ercemapi.2021.17914

Abstract

Dengue is a public health problem in Brazil, cases of the disease started to grow in Paraíba. The epidemiological bulletin of Paraíba, released in August 2021, reports an increase of 53% of cases compared to the previous year. Machine Learning (ML) and Deep Learning techniques are being used as tools to predict the disease and support its combat. Using Random Forest (RF), Support Vector Regression (SVR), Multilayer Perceptron (MLP), Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) techniques, this article presents a system capable of predicting hospital admissions caused by dengue in the cities Bayeux, Cabedelo, João Pessoa and Santa Rita. The system managed to perform forecasts for Bayeux with an error rate of 0.5290, while in Cabedelo the error was 0.92742, João Pessoa 9.55288 and Santa Rita 0.74551.

References

Appice, A., Gel, Y.R., Iliev, I., Lyubchich and V.,Malerba, D. (2020). A Multi-Stage Machine Learning Approach to Predict Dengue Incidence: A Case Study in Mexico. In IEEE Access, v. 8, p. 52713–52725.

Awad, M. and Khanna, R. (2015) Efficient learning machines: Theories, concepts, and applications for engineers and system designers. In Apress Media LLC.

Beserra, Eduardo B., Freitas, Eraldo M. de, Souza, José T. de, Fernandes, Carlos R. M. and Santos, Keliana D. (2009). Ciclo de vida de Aedes (Stegomyia) aegypti (Diptera, Culicidae) em águas com diferentes características. In Iheringia. Série Zoologia.

Carvajal, T.M., Viacrusis, K.M., Hernandez, L.F.T., et al. (2018). Machine learning methods reveal the temporal pattern of dengue incidence using meteorological factors in metropolitan Manila, Philippines. In BMC Infectious Diseases, v. 18, n. 1.

Carvalho, T. M., Tenório, G. L., Figueiredo, K., Vellasco and M., Caarls, W. (2019) Comparison of Machine Learning Models for Total Dengue Cases Prediction.

Doni, A. R. and Sasipraba, T. (2020). Lstm-Rnn Based Approach for Prediction of Dengue Cases in India. In Ingenierie des Systemes d’Information.

Guo, Pi, Liu, Tao, Zhang, Qin, Wang, Li, Xiao, Jianpeng, Zhang, Qingying, Luo, Ganfeng, Li, Zhihao, He, Jianfeng, Zhang, Yonghui and Ma, Wenjun. (2017) Developing a dengue forecast model using machine learning: A case study in China. In PLoS Neglected Tropical Diseases, v. 11, n. 10.

Kirby, Simon, Paramaguru, Kanya and Warren, James. (2015) The Accuracy of NIESR’s GDP Growth Forecasts. In National Institute Economic Review.

Mussumeci, E. and Codeço Coelho, F. (2020). Large-scale multivariate forecasting models for Dengue - LSTM versus random forest regression. Spatial and Spatio-temporal Epidemiology, v. 35.

Norrby, R. (2014) Outlook for a dengue vaccineClinical Microbiology and Infection. Disponível: [link]. Acesso em: 21 set. 2020.

PARAÍBA. (2021) Boletins Epidemiológicos. Disponível em: [link]. Acesso em: 21 jul. 2021.

Pham, Duc Nghia, Aziz, Tarique, Kohan, Ali, Nellis, Syahrul , Jamil, Juraina Binti Abd ,Khoo, Jing Choco ,Lukose, et al. (2018) How to Efficiently Predict Dengue Incidence in Kuala Lumpur. Proceedings - 2018 In 4th International Conference on Advances in Computing, Communication and Automation, ICACCA

Swaminathan, S. and Khanna, N. (2019) Dengue vaccine development: Global and Indian scenarios. In International Journal of Infectious Diseases, v. 84, p. S80–S86.

Xu, Keqiang, Li, Zhichao, Meng, Fengxia, Tu, Taotian, Xu, Lei. et al. (2020). Forecast of dengue cases in 20 chinese cities based on the deep learning method. In International Journal of Environmental Research and Public Health, v. 17, n. 2.

Zhao, K. and Wang, C. (2017). Sales Forecast in E-commerce using Convolutional Neural Network.