Data Mining to Predict COVID-19 Patients’ Recovery on a Balanced Dataset

Abstract

The coronavirus disease , has caused a considerable increase in hospitalizations of people with different symptoms caused by this disease. Currently, the world needs a quick solution to tackle the further spread of COVID-19. Data mining techniques, machine learning and other artificial intelligence techniques can provide a best patient prognosis infected by coronavirus. This paper applies data mining techniques to predict COVID-19 infected patients' recovery using open dataset with day level information on COVID-19 affected cases of China. We also use minority Downsampling technique to balance the classes we have in the dataset and thus demonstrate the importance of balancing the classes to yield better results. Additionally, the pre-processing methods and the prediction performance using evaluation metrics are presented. Logistic Regression, Decision tree, and Neural Network algorithms are applied directly on the dataset using R programming language. Experimental results show that the neural network provides a lower error and increases the classification accuracy significantly compared with other algorithms.

A Deep Learning Model for Coronavirus 2019 Pneumonia Screening

“…Balancing the training set is a simple way to handle this problem in the dataset. To obtain a balanced dataset [7], we propose to use data augmentation [6] only on data "COVID".…”

“…The F1-score is a single metric that uses the harmonic mean to combine recall and precision. As shown in equation (7).…”