Predictive Modeling for Analysis of Coronavirus Symptoms Using Logistic Regression  

作　　者：Anatoli Nachev 

机构地区：[1]Business Information Systems,University of Galway,Galway H91 TK33,Ireland

出　　处：《Journal of Mechanics Engineering and Automation》2023年第4期93-99,共7页机械工程与自动化（英文版）

摘　　要：This paper presents a case study on the IPUMS NHIS database,which provides data from censuses and surveys on the health of the U.S.population,including data related to COVID-19.By addressing gaps in previous studies,we propose a machine learning approach to train predictive models for identifying and measuring factors that affect the severity of COVID-19 symptoms.Our experiments focus on four groups of factors:demographic,socio-economic,health condition,and related to COVID-19 vaccination.By analysing the sensitivity of the variables used to train the models and the VEC(variable effect characteristics)analysis on the variable values,we identify and measure importance of various factors that influence the severity of COVID-19 symptoms.

关 键 词：COVID-19 supervised learning MODELS CLASSIFICATION logistic regression. 

分 类 号：H31[语言文字—英语]