Conditional Variance Estimation (CVE) is a novel sufficient dimension reduction (SDR) method for additive error regressions with continuous predictors and link function. It operates under the assumption that the predictors can be replaced by a lower dimensional projection without loss of information. Conditional Variance Estimation is fully data driven, does not require the restrictive linearity and constant variance conditions, and is not based on inverse regression as the majority of moment and likelihood based sufficient dimension reduction methods. CVE is shown to be consistent and its objective function to be uniformly convergent. CVE outperforms the mean average variance estimation, (MAVE), its main competitor, in several simulation settings, remains on par under others, while it always outperforms inverse regression based linear SDR methods, such as Sliced Inverse Regression.
en
dc.language.iso
en
-
dc.relation.ispartof
Bernoulli
-
dc.subject
Regression
en
dc.subject
Mean subspace
en
dc.subject
Nonparametric
en
dc.subject
Statistics and Probability
en
dc.subject
Minimum average variance estimation
en
dc.subject
Dimension reduction
en
dc.title
Conditional variance estimator for sufficient dimension reduction
en
dc.type
Artikel
de
dc.type
Article
en
dc.description.startpage
1862
-
dc.description.endpage
1891
-
dc.type.category
Original Research Article
-
tuw.container.volume
28
-
tuw.container.issue
3
-
tuw.journal.peerreviewed
true
-
tuw.peerreviewed
true
-
tuw.researchTopic.id
A4
-
tuw.researchTopic.id
C6
-
tuw.researchTopic.name
Mathematical Methods in Economics
-
tuw.researchTopic.name
Modelling and Simulation
-
tuw.researchTopic.value
75
-
tuw.researchTopic.value
25
-
dcterms.isPartOf.title
Bernoulli
-
tuw.publication.orgunit
E105-08 - Forschungsbereich Angewandte Statistik
-
tuw.publisher.doi
10.3150/21-bej1402
-
dc.identifier.eissn
1573-9759
-
dc.description.numberOfPages
30
-
wb.sci
true
-
wb.sciencebranch
Mathematik
-
wb.sciencebranch.oefos
1010
-
wb.facultyfocus
Wirtschaftsmathematik und Stochastik
de
wb.facultyfocus
Mathematical Methods in Economics and Stochastics
en
wb.facultyfocus.faculty
E100
-
item.languageiso639-1
en
-
item.fulltext
no Fulltext
-
item.openairetype
research article
-
item.grantfulltext
none
-
item.openairecristype
http://purl.org/coar/resource_type/c_2df8fbb1
-
item.cerifentitytype
Publications
-
crisitem.author.dept
E105-08 - Forschungsbereich Angewandte Statistik
-
crisitem.author.dept
E105-08 - Forschungsbereich Angewandte Statistik
-
crisitem.author.parentorg
E105 - Institut für Stochastik und Wirtschaftsmathematik
-
crisitem.author.parentorg
E105 - Institut für Stochastik und Wirtschaftsmathematik