Predicting H1N1 and Seasonal Flu : Vaccine Cases using Ensemble Learning approach

Abstract

We propose a data-driven machine learning model to predict the likelihood of a person being vaccinated against H1N1 and seasonal flu. In 2009, a pandemic that was caused by the H1N1 influenza virus, named "swine flu", spread like wildfire across the world. Researchers and Healthcare specialists estimated that in the initial year, it was responsible for between \underline{150, 000\ {\text{to}}\ 600, 000\ {\text{deaths}}} in the whole world. A vaccine against the H1N1 (swine flu) virus was made available in October 2009. We have implemented 9 machine learning models. The models include MlBox (model1), TPOT (model 2), Random forest (model 3), MLP (model 4), Linear regression, (model 5), Decision trees (model 6), polynomial feature (model 7), XgBoost (model 8) and CatBoost (model 9). Out of these 9 models, CatBoost gave the best performance with an accuracy of 0.8617 followed by the XgBoost and MlBox. Hence, this study showed that cast boost gave the best performance of all the models and can be used for further prediction models in this category. © 2020 IEEE.