On marginal sliced inverse regression for ultrahigh dimensional model-free feature selection (Q510686): Difference between revisions
From MaRDI portal
Created a new Item |
Set OpenAlex properties. |
||
(4 intermediate revisions by 4 users not shown) | |||
Property / author | |||
Property / author: Zhou Yu / rank | |||
Property / author | |||
Property / author: Jun Shao / rank | |||
Property / author | |||
Property / author: Zhou Yu / rank | |||
Normal rank | |||
Property / author | |||
Property / author: Jun Shao / rank | |||
Normal rank | |||
Property / review text | |||
This paper aims to develop selection consistency for ultrahigh dimensional settings. An approach called marginal sliced inverse regression (SIR) is proposed for model-free feature screening and variable selection. Two distinct procedures, Dantzig selector and sparse precision matrix estimation, are incorporated to get two versions of sample level marginal SIR utilities. Both procedures lead to model-free variable selection consistency with predictor dimensionality \(p\) diverging at an exponential rate of the sample size \(n\). As a special case of marginal SIR, the correlation among the predictors is ignored and marginal independence SIR is proposed. Marginal independence SIR is closely related to many existing independence screening procedures in the literature, and achieves model-free screening consistency in the ultrahigh dimensional setting. The finite sample performances of the proposed procedures are studied through synthetic examples and an application to the small round blue cell tumour data is given. | |||
Property / review text: This paper aims to develop selection consistency for ultrahigh dimensional settings. An approach called marginal sliced inverse regression (SIR) is proposed for model-free feature screening and variable selection. Two distinct procedures, Dantzig selector and sparse precision matrix estimation, are incorporated to get two versions of sample level marginal SIR utilities. Both procedures lead to model-free variable selection consistency with predictor dimensionality \(p\) diverging at an exponential rate of the sample size \(n\). As a special case of marginal SIR, the correlation among the predictors is ignored and marginal independence SIR is proposed. Marginal independence SIR is closely related to many existing independence screening procedures in the literature, and achieves model-free screening consistency in the ultrahigh dimensional setting. The finite sample performances of the proposed procedures are studied through synthetic examples and an application to the small round blue cell tumour data is given. / rank | |||
Normal rank | |||
Property / reviewed by | |||
Property / reviewed by: Yuehua Wu / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 62H20 / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 62G08 / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 62H99 / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 62J99 / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 62P10 / rank | |||
Normal rank | |||
Property / zbMATH DE Number | |||
Property / zbMATH DE Number: 6684021 / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
marginal coordinate test | |||
Property / zbMATH Keywords: marginal coordinate test / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
sliced inverse regression | |||
Property / zbMATH Keywords: sliced inverse regression / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
sufficient dimension reduction | |||
Property / zbMATH Keywords: sufficient dimension reduction / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
sure independence screening | |||
Property / zbMATH Keywords: sure independence screening / rank | |||
Normal rank | |||
Property / MaRDI profile type | |||
Property / MaRDI profile type: MaRDI publication profile / rank | |||
Normal rank | |||
Property / full work available at URL | |||
Property / full work available at URL: https://doi.org/10.1214/15-aos1424 / rank | |||
Normal rank | |||
Property / OpenAlex ID | |||
Property / OpenAlex ID: W2551812373 / rank | |||
Normal rank | |||
links / mardi / name | links / mardi / name | ||
Latest revision as of 11:54, 30 July 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | On marginal sliced inverse regression for ultrahigh dimensional model-free feature selection |
scientific article |
Statements
On marginal sliced inverse regression for ultrahigh dimensional model-free feature selection (English)
0 references
13 February 2017
0 references
This paper aims to develop selection consistency for ultrahigh dimensional settings. An approach called marginal sliced inverse regression (SIR) is proposed for model-free feature screening and variable selection. Two distinct procedures, Dantzig selector and sparse precision matrix estimation, are incorporated to get two versions of sample level marginal SIR utilities. Both procedures lead to model-free variable selection consistency with predictor dimensionality \(p\) diverging at an exponential rate of the sample size \(n\). As a special case of marginal SIR, the correlation among the predictors is ignored and marginal independence SIR is proposed. Marginal independence SIR is closely related to many existing independence screening procedures in the literature, and achieves model-free screening consistency in the ultrahigh dimensional setting. The finite sample performances of the proposed procedures are studied through synthetic examples and an application to the small round blue cell tumour data is given.
0 references
marginal coordinate test
0 references
sliced inverse regression
0 references
sufficient dimension reduction
0 references
sure independence screening
0 references