Download PDFOpen PDF in browser

Prediction of Covid-19 Severity Level Using XGBoost Algorithm: a Machine Learning Approach Based on SIR Epidemical Model

EasyChair Preprint 6097

8 pagesDate: July 16, 2021

Abstract

Covid-19 formally termed as "2019 novel coronavirus", is disrupting socio-economic conditions throughout the world. Due to the unavailability of efficient ways to predict the severity level of Covid-19, governmental officials and policymakers of different countries are facing difficulties to take precautionary measures for minimizing risks. This paper presents a model trained to predict Covid-19 situation severity level using XGBoost which is a gradient boosting algorithm. To categorize severity level, SIR epidemiological method has been employed which can express the current condition of any area affected by contagious diseases like Covid-19 analyzing the number of susceptible people which stands for S, the number of infected people stands for I, and the number of recovered people stands for R. By comparing the evaluation metrics of our model with other models based on different machine learning algorithms, it is deduced that the model performs better for less training time(speed), better accuracy rate, and has the ability to reduce over-fitting.

Keyphrases: COVID-19, SIR, XGBoost, machine learning, prediction

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@booklet{EasyChair:6097,
  author    = {Labeba Tahsin and Shaily Roy},
  title     = {Prediction of Covid-19 Severity Level Using XGBoost Algorithm: a Machine Learning Approach Based on SIR Epidemical Model},
  howpublished = {EasyChair Preprint 6097},
  year      = {EasyChair, 2021}}
Download PDFOpen PDF in browser