# Analysis of the Effectiveness of Face-Coverings on the Death Ratio of COVID-19 Using Machine Learning

**Authors:** Ali Lafzi, Miad Boodaghi, Siavash Zamani, Niyousha Mohammadshafie, Veeraraghava Raju Hasti  
**arXiv:** [2102.04419](https://arxiv.org/abs/2102.04419)  
**Date:** 2021-02-08  
**Summary page:** https://www.summarizepaper.com/en/paper/2102.04419v2/

## Key points

- The study aims to expand and refine the existing summary related to the recent outbreak of COVID-19.
- Survey data from the New York Times available on GitHub was used to quantify people's adherence to the mask mandate order issued by state governors in the US.
- The average death ratio was defined as a measure of disease severity.
- Machine learning classification algorithms were used to investigate how changes in mortality rate correlate with features reflecting adherence to the mask mandate order in different counties.
- Data from one month before and after executive orders were signed for each preventive measure was collected from USAFACTS website for three West Coast states: California, Oregon, and Washington.
- Out of 130 counties, there were 47 instances where death ratios decreased after implementing mask mandates, 30 instances where they increased, and 53 instances where no change occurred.
- Machine learning algorithms were employed for training and prediction.
- The study aims to examine the correlation between public adherence to mask mandates and changes in mortality rate using machine learning classification algorithms.
- Data analysis focuses on three West Coast states during a specific time period to evaluate the effectiveness of mask mandates as a preventive measure against COVID-19.

## Summary

In this study, we aim to expand and refine the existing summary by incorporating additional context and details related to the recent outbreak of COVID-19. To quantify people's adherence to the mask mandate order issued by state governors in the US, we utilized survey data provided by the New York Times available on GitHub. We defined a parameter called the average death ratio which serves as a measure of disease severity. Using different machine learning classification algorithms, we sought to investigate how changes in mortality rate correlate with certain features chosen to reflect adherence to the mask mandate order in different counties. To obtain unbiased data, information from one month before and after executive orders were signed for each preventive measure was collected from USAFACTS website for three West Coast states: California (June 18th, 2020), Oregon (June 19th, 2020), and Washington (June 26th, 2020). Out of a total sample size of 130 counties, there were 47 instances where death ratios decreased after implementing mask mandates, 30 instances where they increased, and 53 instances where no change occurred. The methodology employed machine learning algorithms for training and prediction. In conclusion, this study aims to examine the correlation between public adherence to mask mandates and changes in mortality rate using machine learning classification algorithms. By analyzing data from three West Coast states specifically during this time period we hope to shed light on the effectiveness of mask mandates as a preventive measure against COVID-19.