Advances in variable selection methods I: Causal selection methods versus stepwise regression and principal component analysis on data of known and unknown functional relationships

dc.contributor.author	Ssegane, H	en
dc.contributor.author	Tollner, EW	en
dc.contributor.author	Mohamoud, YM	en
dc.contributor.author	Rasmussen, TC	en
dc.contributor.author	Dowd, JF	en
dc.date.accessioned	2014-06-06T06:51:36Z
dc.date.available	2014-06-06T06:51:36Z
dc.date.issued	2012	en
dc.identifier.issn	00221694	en
dc.identifier.uri	http://dx.doi.org/10.1016/j.jhydrol.2012.01.008	en
dc.identifier.uri	http://62.217.125.90/xmlui/handle/123456789/5594
dc.subject	Causal variable selection	en
dc.subject	Principal component analysis	en
dc.subject	Stepwise regression	en
dc.subject	Streamflow percentiles	en
dc.subject	Watershed variables	en
dc.subject.other	Descriptors	en
dc.subject.other	First-order	en
dc.subject.other	Functional relationship	en
dc.subject.other	Hillslopes	en
dc.subject.other	Hydrological prediction	en
dc.subject.other	Hydrological process	en
dc.subject.other	Markov Blankets	en
dc.subject.other	Markov boundary	en
dc.subject.other	Predictive power	en
dc.subject.other	Principal components analysis	en
dc.subject.other	Selection methods	en
dc.subject.other	Stepwise regression	en
dc.subject.other	Upscaling	en
dc.subject.other	Variable selection	en
dc.subject.other	Variable selection methods	en
dc.subject.other	Watershed characteristics	en
dc.subject.other	Watershed scale	en
dc.subject.other	Landforms	en
dc.subject.other	Principal component analysis	en
dc.subject.other	Regression analysis	en
dc.subject.other	Stream flow	en
dc.subject.other	Watersheds	en
dc.subject.other	accuracy assessment	en
dc.subject.other	data set	en
dc.subject.other	hillslope	en
dc.subject.other	hydrological change	en
dc.subject.other	Markov chain	en
dc.subject.other	prediction	en
dc.subject.other	principal component analysis	en
dc.subject.other	regression analysis	en
dc.subject.other	runoff	en
dc.subject.other	streamflow	en
dc.subject.other	watershed	en
dc.title	Advances in variable selection methods I: Causal selection methods versus stepwise regression and principal component analysis on data of known and unknown functional relationships	en
heal.type	journalArticle	en
heal.identifier.primary	10.1016/j.jhydrol.2012.01.008	en
heal.publicationDate	2012	en
heal.abstract	Hydrological predictions at a watershed scale are commonly based on extrapolation and upscaling of hydrological behavior at plot and hillslope scales. Yet, dominant hydrological drivers at a hillslope may not be as dominant at the watershed scale because of the heterogeneity of watershed characteristics. With the availability of quantifiable watershed data (watershed descriptors and streamflow indices), variable selection can provide insight into the dominant watershed descriptors that drive different streamflow regimes. Stepwise regression and principal components analysis have long been used to select descriptive variables for relating runoff to climate and watershed descriptors. Questions have remained regarding the robustness of the selected descriptors. This paper evaluates five new approaches: Grow-Shrink, GS; a variant of Incremental Association Markov Boundary, interIAMBnPC; Local Causal Discovery, LCD2; HITON Markov Blanket, HITON-MB; and First-Order Utility, FOU. We demonstrate their performance by quantifying their accuracy, consistency and predictive potential compared to stepwise regression and principal component analysis on two known functional relationships. The results show that the variables selected by HITON-MB and the first-order utility are the most accurate while variables selected by Stepwise regression, although not accurate have a high predictive potential. Therefore, a model with high predictive power may not necessary represent the underlying hydrological processes of a watershed system. © 2012 Elsevier B.V.	en
heal.journalName	Journal of Hydrology	en
dc.identifier.volume	438-439	en
dc.identifier.doi	10.1016/j.jhydrol.2012.01.008	en
dc.identifier.spage	16	en
dc.identifier.epage	25	en