e-space
Manchester Metropolitan University's Research Repository

COVID-19 patient health prediction using boosted random forest algorithm

Iwendi, C and Bashir, AK and Peshkar, A and Sujatha, R and Chatterjee, JM and Pasupuleti, S and Mishra, R and Pillai, S and Jo, O (2020) COVID-19 patient health prediction using boosted random forest algorithm. Frontiers in Public Health, 8.

[img]
Preview

Available under License Creative Commons Attribution.

Download (1MB) | Preview

Abstract

© 2020 Iwendi, Bashir, Peshkar, Sujatha, Chatterjee, Pasupuleti, Mishra, Pillai and Jo. Integration of artificial intelligence (AI) techniques in wireless infrastructure, real-time collection, and processing of end-user devices is now in high demand. It is now superlative to use AI to detect and predict pandemics of a colossal nature. The Coronavirus disease 2019 (COVID-19) pandemic, which originated in Wuhan China, has had disastrous effects on the global community and has overburdened advanced healthcare systems throughout the world. Globally; over 4,063,525 confirmed cases and 282,244 deaths have been recorded as of 11th May 2020, according to the European Centre for Disease Prevention and Control agency. However, the current rapid and exponential rise in the number of patients has necessitated efficient and quick prediction of the possible outcome of an infected patient for appropriate treatment using AI techniques. This paper proposes a fine-tuned Random Forest model boosted by the AdaBoost algorithm. The model uses the COVID-19 patient’s geographical, travel, health, and demographic data to predict the severity of the case and the possible outcome, recovery, or death. The model has an accuracy of 94% and a F1 Score of 0.86 on the dataset used. The data analysis reveals a positive correlation between patients’ gender and deaths, and also indicates that the majority of patients are aged between 20 and 70 years.

Impact and Reach

Statistics

Downloads
Activity Overview
14Downloads
54Hits

Additional statistics for this dataset are available via IRStats2.

Altmetric

Actions (login required)

View Item View Item