A Deep learning approach for prediction of COVID-19

Due to the crisis of human health, the medical system supports artificial intelligence like technologies to control and monitor the spread of pandemic COVID-19. Moreover, we focus on coronavirus spread prediction and feature analysis by using deep learning techniques. Considering massive dataset, it makes sense to utilize mathematical models and deep learning tools to predict the affected area and it is also helpful to predict the disease rate in near future. In this paper, we want to make an efficient prediction algorithm. Firstly, the SMOTE-ENN tool is used to handle the imbalance dataset then XGBoost and sprace autoencoder techniques are used for feature selection as well as for classification. To know the importance of the feature SHAPLY technique is used to analyse the data and explore the impact of data set features.