A critical assessment of shrinkage-based regression approaches for estimating the adverse health effects of multiple air pollutants |
| |
| Institution: | 1. School of Geosciences and Info-Physics, Central South University, Changsha, Hunan 410083, China;2. School of Geography and Planning, Sun Yat-Sen University, Guangzhou, Guangdong 510275, China;3. School of Geography and Environment, Oxford University, Oxford, UK;4. National Geographic Conditions Monitoring Research Center, Chinese Academy of Surveying and Mapping, Beijing 100830, China;1. School of Energy and Environmental Engineering, University of Science and Technology Beijing, Beijing 100083, China;2. Department of Environmental Health Sciences, School of Public Health, University at Albany, State University of New York, Albany, NY 12144, USA;3. Institute of Reproductive and Child Health/Ministry of Health Key Laboratory of Reproductive Health and Department of Epidemiology and Biostatistics, School of Public Health, Peking University, Beijing 100083, China;4. National Institute for Nutrition and Health, Chinese Center for Disease Control and Prevention, Beijing 100050, China;5. State Key Joint Laboratory of Environmental Simulation & Pollution Control, School of Environment, Beijing Normal University, Beijing 100875, China;6. Center for Environment, Energy, and Economy, Harrisburg University, Harrisburg PA17101, USA;7. Nicholas School of the Environment and Global Health Institute, Duke University, Durham, NC, USA;8. Duke Kunshan University, Kunshan, Jiangsu Province, China;1. Foundation for Research and Technology — Hellas, Institute of Applied and Computational Mathematics, N. Plastira 100, Vassilika Vouton, 70013 Heraklion, Greece;2. School of Mathematical & Statistical Sciences, Arizona State University, Tempe, AZ 85287, USA |
| |
| Abstract: | Most investigations of the adverse health effects of multiple air pollutants analyse the time series involved by simultaneously entering the multiple pollutants into a Poisson log-linear model. Concerns have been raised about this type of analysis, and it has been stated that new methodology or models should be developed for investigating the adverse health effects of multiple air pollutants. In this paper, we introduce the use of the lasso for this purpose and compare its statistical properties to those of ridge regression and the Poisson log-linear model. Ridge regression has been used in time series analyses on the adverse health effects of multiple air pollutants but its properties for this purpose have not been investigated. A series of simulation studies was used to compare the performance of the lasso, ridge regression, and the Poisson log-linear model. In these simulations, realistic mortality time series were generated with known air pollution mortality effects permitting the performance of the three models to be compared. Both the lasso and ridge regression produced more accurate estimates of the adverse health effects of the multiple air pollutants than those produced using the Poisson log-linear model. This increase in accuracy came at the expense of increased bias. Ridge regression produced more accurate estimates than the lasso, but the lasso produced more interpretable models. The lasso and ridge regression offer a flexible way of obtaining more accurate estimation of pollutant effects than that provided by the standard Poisson log-linear model. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
