Abstract
Presented at MOBIMEDIA Conference
Particulate matter 2.5 (PM2.5) are fine particles can penetrate deeply into our lungs and other airways areas because of their small sizes. Sometimes these fine particles may even enter the bloodstreams. Only a few researches studied the relation between PM2.5 and lung cancers. In this paper, innovative machine learning and spatiotemporal interpolation methods were used to compute historical PM2.5 interpolation data in the contiguous United States. Time series analysis (including seasonal ARIMA models, lagged regressions, generalized estimating equations) is then applied to lung and bronchial cancers and PM2.5 data. Based on our current data covering a 15-year span (1999-2014), PM2.5 doesn’t have a strong effect on lung and bronchial cancer rates in the United States at either the national or state level. However, the most urban state, New Jersey, and highest PM2.5 state, California, have a relatively greater tendency to have significant PM2.5 effect among all contiguous U.S. states.
Original language | American English |
---|---|
State | Published - Jun 1 2018 |
Event | MOBIMEDIA 2018 Conference - Duration: Jun 1 2018 → … |
Conference
Conference | MOBIMEDIA 2018 Conference |
---|---|
Period | 06/1/18 → … |
Keywords
- PM2.5
- lung and bronchial cancers
- spatiotemporal interpolation
- time series analysis
- regression analysis
DC Disciplines
- Biostatistics
- Environmental Public Health
- Epidemiology
- Public Health