Journal of Data Science ›› 2020, Vol. 18 ›› Issue (3): 483-494.doi: 10.6339/JDS.202007_18(3).0015

Previous Articles     Next Articles

COVID-19 Fatality: A Cross-Sectional Study using Adaptive Lasso Penalized Sliced Inverse Regression

Kaida Cai2, Wenqing He1, and Grace Y. Yi1,2

  1. 1 Department of Statistical and Actuarial Sciences, University of Western Ontario, London, Ontario, Canada
    2 Department of Computer Science, University of Western Ontario, London, Ontario, Canada
  • Online:2020-07-21 Published:2020-07-22

Abstract: Coronavirus disease 2019 (COVID-19) is an infectious disease caused by severe acute respiratory syndrome coronvirus, which was declared as a global pandemic by the World Health Organi- zation on March 11, 2020. In this work, we conduct a cross-sectional study to investigate how the infection fatality rate (IFR) of COVID-19 may be associated with possible geographical or demographical features of the infected population. We employ a multiple index model in combi- nation with sliced inverse regression to facilitate the relationship between the IFR and possible risk factors. To select associated features for the infection fatality rate, we utilize an adaptive Lasso penalized sliced inverse regression method, which achieves variable selection and sufficient dimension reduction simultaneously with unimportant features removed automatically. We ap- ply the proposed method to conduct a cross-sectional study for the COVID-19 data obtained from two time points of the outbreak.

Key words: coronavirus disease 2019, infection fatality rate, multiple index model, risk factors