Data-Driven Model For Influenza Prediction Incorporating Environmental Effects

Influenza is one of the most severe and prevalent epidemic that causes mortality and morbidity. The researcher focused on early forecasting to prevent and control the outbreak of the flu disease, which it may reduce their impact on our daily lives. We propose a model based on machine learning methods that is capable of making timely influenza prediction using the impact of many environmental factors such as climatic variables, air pollutants and geographical proximity. Our significant contribution is to incorporate the impact of this environmental factors changes on the spread of the disease with a machine learning method to improve the performance of the influenza prediction models. We use multiple data sources including Illness Like Influenza (ILI) data, climatic factors, air pollutant and geographic proximity that have significant correlation with ILI rate. In this paper, we compare the proposed model with two methods and with the actual value to prove the effectiveness of our approach.