Efficient Dengue Spread Prediction Using Machine Learning Models with Various Preprocessing Techniques

The viral disease dengue, spread by mosquitoes, is a severe hazard to public health worldwide. Predicting dengue outbreaks is an essential field of study that aims to stop and lessen the effects of the virus spread by mosquitoes. Utilizing modern technologies like data analytics and ML(Machine Learning) will significantly improve the capacity to predict and contain dengue outbreaks. The work examines how the effectiveness of Convolutional Neural Networks (CNNs) in forecasting and evaluating Dengue outbreaks is affected by several scaling strategies, including Min-Max Scaling, Z-Score Standardization, and Robust Scaling. For disease control to be effective, outbreaks must be predicted with accuracy and timeliness. Before feeding the input data into the CNN architecture, preprocess takes place in experiments utilizing Robust Scaling, Z-Score Standardization, and Min-Max Scaling. The model is methodically assessed using necessary measures like recall, accuracy, and precision. The findings show that robust scaling consistently performs better than the other scaling methods, improving CNN’s capacity to identify characteristics and patterns important for predicting Dengue outbreaks. The result was that Robust Scaling+CNN gave the best results, with an accuracy of $93 \%$, precision, and recall of 0.92 and 0.85, respectively. The tool used for execution is Python.