Proper orthogonal decomposition methods for the analysis of real-time data: Exploring peak clustering in a secondhand smoke exposure intervention
•A POD/k-means clustering algorithm classified time-series peaks into one of two empirically defined classes.•A behavioral intervention increased the proportion of peaks from the class associated with shorter durations.•Distinguishing physical properties of the peaks were identified with >60% acc...
Saved in:
Published in | Journal of computational science Vol. 11; pp. 102 - 111 |
---|---|
Main Authors | , , , , , , , |
Format | Journal Article |
Language | English |
Published |
Netherlands
Elsevier B.V
01.11.2015
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | •A POD/k-means clustering algorithm classified time-series peaks into one of two empirically defined classes.•A behavioral intervention increased the proportion of peaks from the class associated with shorter durations.•Distinguishing physical properties of the peaks were identified with >60% accuracy.
This work explores a method for classifying peaks appearing within a data-intensive time-series. We summarize a case study from a clinical trial aimed at reducing secondhand smoke exposure via the installation of air particle monitors in households. Proper orthogonal decomposition (POD) in conjunction with a k-means clustering algorithm assigns each data peak to one of two clusters. Aversive feedback from the monitors increased the proportion of short-duration, attenuated peaks from 38.8% to 96.6%. For each cluster, a distribution of parameters from a physics-based model of airborne particles is estimated. Peaks generated from these distributions are correctly identified by POD/clustering with >60% accuracy. |
---|---|
Bibliography: | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
ISSN: | 1877-7503 1877-7511 |
DOI: | 10.1016/j.jocs.2015.10.006 |