A Comparative Approach To Predict Corona Virus Using Machine Learning

Abstract

Coronavirus disease (COVID-19), is one of the most infectious diseases which reshaped our everyday lives globally in the 21 st century. Technology progressions have a rapid effect on every field of life, be it the medical domain or any other. More than 250 countries have been affected by COVID in no matter of time. The Indian government is making the necessary steps to control the spread of virus in the society. People all over the world are vulnerable to its consequences in the future. In a pandemic like this, people often worry whether they show a symptom of COVID-19 or not. Various AI methods have been applied successfully in epidemic studies. This paper presents the prediction and analysis of COVID-19 using various machine learning algorithms. In the present study, ML-based enhanced model is implemented to predict the possible threat of COVID-19 all over the world and the algorithms used in these models classifies the COVID patients based on several subsets of features and predicts their likeliness to get affected to this disease. This model uses 20 metrics including the patient's geographical location, travel history, health record statistics, etc., to predict the severity of the case and the feasible outcome. This research finds the patients exposed to Covid-19 and could be used as a reference, by the patients before consulting further with the doctor. The model developed using K-Nearest Neighbors (KNN) is effective with a prediction accuracy of 98.34%, Recall of 97% and an F1-S core of 0.97. Overall, this paper proposes a simple and practicable method to quickly identify and predict the high risk patients and provide priority to them for treatment so that the fatality rate can be decreased.

Automatic COVID-19 Prediction with Comprehensible Machine Learning Models

“…This visualization further demonstrates the model's ability to analyse input data and generate predictions, highlighting its potential to assist healthcare professionals in decision-making processes related to COVID-19 diagnosis and management. Age has been identified as one of the most important characteristics that contribute to the prediction of case severity in research on mortality or severity prediction [20][21][22][23]. Age was one of the top 10 features in our analysis out of the 25 features we utilized in our prediction model.…”

Assessment of SQL and NoSQL Systems to Store and Mine COVID-19 Data

“…worldwide. As a result, multiple papers have been published regarding this issue [5,6,[22][23][24][25][26][27]. In addition, a significant amount of research has been put forward using artificial intelligence on COVID-19 data.…”

“…In the studies [5,[22][23][24] Decision Trees were the best algorithm in the prediction models with accuracies of 99.85%, 94.99%, 98.7%, and 99.98%, respectively. In [6,[25][26][27], different algorithms such as K-Nearest Neighbor, Logistic Regression, Bayesian Network, and Random Forest proved to be the best in prediction models with accuracies of 98%, 87%, 89.31%, and 90.83%, respectively. These studies were beneficial for combat against COVID-19, with the accuracies reaching high values-between 99.98% [24] and 87% [25].…”

“…Orange Data Mining provides a vast Data Mining algorithm's range. Naïve Bayes, Decision Tree, Random Forest, and Logistic Regression were used in this study as state-ofthe-art [5,6,[22][23][24][25][26][27]. The following subsections briefly describe each of these algorithms.…”

“…Healthcare organizations are currently capable of collecting a large amount of data that can be used to generate helpful insights once mined. Data Mining is an advanced artificial intelligence technique used for discovering novel, useful, and valid hidden patterns or knowledge from datasets [5,6]. The acquired knowledge through this process has been improving work efficiency and strengthening decision-making.…”