Seasonal Infectious Disease Prediction based on Electronic Patient Health Records using Boosted Random Forest Algorithms

ABSTRACT

Environmental condition and climatic factors are important elements of infection outbreaks. Some illness can change their effect of spread based on the weather condition. Such as Dengue increase their transmission in monsoon and winter seasons. Because these are rainy periods. So the wet occasions can provoke the dengue mosquito production. Meteorological changes play an essential aspect to trigger a new sickness. And also it produces new pathogens. Machine learning is an efficient component to predict seasonal wise infectious disease depends on climatic changes. In this paper, the researcher discusses three pandemics such as COVID-19, Dengue and Flu. As well as how these three epidemics can modify their dissemination based upon the weather pattern. Furthermore, the researcher predicts which season influences which ailment based on the patient health records. And also discuss various disease prediction algorithms such as Naive bayes, Decision tree, KNN (K-nearest neighbors), Boosted random forest and SVM (Support vector machine). The Boosted random Forest algorithm gives the 95% accuracy for forecasting the seasonal infections based on the patients electronic health record. By using Boosted Random Forest algorithm the researcher finds the winter season is the most suitable season produces high aliments compared to other seasons. It is very helpful to the health analyst to identify the seasonal diseases © 2022 IEEE.