A Deep Learning Approach for Dengue Fever Prediction in Malaysia Using LSTM with Spatial Attention

Abstract

Dengue is a very common disease typically found in a widespread hot region. It is a deadly disease caused by female Aedes mosquitoes. Patients who have diagnosed with different types of dengue needed different types of treatment. Many experts are experimenting to recognize and exploring some of the new features on dengue disease from last few years. In this research paper, we have predicted dengue fever by implementing various machine learning algorithms such as two boosted decision tree, two-class Bayes point machine, multiclass decision forest, and boosted decision tree regression in a dataset. This dataset is collected from our generated survey of different people who are currently affected by dengue fever or already suffered from dengue fever. We also used tenfold cross-validation to estimate the performance of our machine learning model. We have also used Azure machine learning studio to predict and evaluate data, and we also compare the performance of all techniques that we have used. We also showed the accuracy of all the classification and regression technique. This research paper is the first unique paper based on Bangladesh region which will be used to detecting the dengue fever with 95% accuracy. People will be aware of this dangerous disease and can take necessary actions. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.