Prediction of Dengue Fever Outbreak Based on Climate Factors Using Fuzzy-Logistic Regression

摘要

Dengue fever outbreak prediction is said to be one way that can be used to restrain the spread of dengue fever. Thus, the accuracy of the outbreak prediction system becomes essential. Furthermore, the factors involved in the prediction are also crucial to note. This study combines temperature, rainfall, humidity, wind speed, and the number of dengue cases to predict the outbreak of dengue fever. The fuzzy-logistic regression model is used based on its compatibility with the input and output characteristics. The result shows that the fuzzy-logistic regression model can produce outbreak predictions for validation data in other regions with an average performance of 79.93%. This average performance is 14.95% higher than the average accuracy of the Neural Network, Random Forest, and Naive Bayes approaches. The prediction results for the next 24 periods show that the outbreak will occur seven times. Dengue fever case and temperature are two variables that have more influence than other variables.