Influenza-type illnesses and air pollutants of particulate matter < 2.5μm (PM2.5): an application of Archetypal Analysis to identify spatiotemporal structure

Document Type

Presentation Abstract

Presentation Date

2-4-2019

Abstract

Particulate matter (PM2.5) readings are often included in air quality reports from environmental authorities as it can pose the most danger when it builds up in human respiratory system and increases the risk of respiratory infections and lung diseases. Understanding the spatio-temporal variability of upper respiratory illness and its dependence upon air quality in Montana is an area of active research in the public health sphere.

Archetypal analysis (AA), Culter and Breiman 1994, is introduced as a method to decompose and characterize structures within spatio-temporal data. AA seeks to synthesize a set of multivariate observations through a few, not necessarily observed points (archetypes), which lie on the boundary of the data cloud. This method is new in climate science, although it has been around for more than two decades in pattern recognition.

The goal of this presentation is to examine the spatio-temporal variability of two sets of weekly influenza cases and PM2.5 across Montana between 2008-2018 through AA. Compared to other conventional methods, such as PCA, the results provide the direct link to the observations which facilitate the interpretation. The patterns exposed by AA in both cases are contrasted, as one data set is approximately spatially continuous (PM) and the other is not (Flu counts).

Additional Details

Monday, February 4, 2019 at 3:00 p.m. in Math 103
Refreshments at 4:00 p.m. in Math Lounge 109

This document is currently not available here.

Share

COinS