Source contributions to ambient aerosol calculated by discriminat partial least squares regression (PLS)

Abstract

Partial least squares regression (PLS) is proposed for solving air pollution source apportionment problems as an alternative method to the frequently used chemical mass balance technique. A discriminant PLS is used to calculate linear mixing proportions for a synthetic ambient aerosol data set where the truth is known. Without sacrificing orthogonality of the source profiles, PLS can resolve the emission sources and accurately predict the emission source contributions. Further extensions of the PLS approach to environmental receptor modelling are discussed. KEY WORDS Partial least squares Receptor modelling Colinearities SNTNF Senior Visiting Scientist. Current address: EPA Environmental Research Lab., 200 S.W. 35th Street, Corvallis, OR 97333, U.S.A. 0886-9383/88/04028 1-16$08 .OO

Metabolic evidence for biogeographic isolation of the extremophilic bacterium Salinibacter ruber

“…The statistical model used was partial least squares for discriminant analysis (PLS-DA) (Sjö ström et al, 1986;Stahle and Wold, 1987;Vong et al, 1988;Kemsley, 1996). PLS-DA is a regression extension of the principal component analysis .…”

Diagnosis of Dementias Using Partial Least Squares Discriminant Analysis

“…It is important in this context to emphasize that the subgroup ing per se was not manipulated by the use of PLS (i.e. use of previous clinical diagnoses) but only the point within the 29-dimensional descriptor space from which we chose to view our patients [19]. Thus, we are convinced that subgrouping within consecutive dementia patients, as demonstrated in the present study, was genuine and, in addition, it was evident despite the deliberate removal of any variable containing information of age or age at onset of disease.…”

Multi-Label Dimensionality Reduction

“…Since PLS and CCA share many common features, PLS is also used widely in many applications as a classification tool [145,179], although it is not inherently designed for this purpose. Some typical applications of PLS in classification include Alzheimer's disease discrimination [180], classification between Arabica and Robusta coffee beans [181], water pollution classification [182], separation between active and inactive compounds in a quantitative structure-activity relationship study [183], hard red wheat classification [184], microarray classification [160], and other applications in a variety of other areas [185][186][187][188].…”