The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Pemilihan Ciri tanpa pengawasan ialah teknik pengurangan dimensi yang penting untuk mengatasi data berdimensi tinggi. Ia tidak memerlukan maklumat label terdahulu, dan baru-baru ini telah menarik banyak perhatian. Walau bagaimanapun, ia tidak boleh menggunakan sepenuhnya maklumat diskriminasi sampel, yang mungkin menjejaskan prestasi pemilihan ciri. Untuk menangani masalah ini, dalam surat ini, kami mencadangkan kaedah regresi label maya diskriminatif novel (DVLR) untuk pemilihan ciri tanpa pengawasan. Dalam DVLR, kami membangunkan fungsi regresi label maya untuk membimbing pemilihan ciri berasaskan pembelajaran subruang, yang boleh memilih lebih banyak ciri diskriminatif. Selain itu, istilah analisis diskriminasi linear (LDA) digunakan untuk menjadikan model lebih diskriminatif. Untuk menjadikan model lebih mantap dan memilih ciri yang lebih representatif, kami mengenakannya ℓ2,1-norma pada regresi dan syarat pemilihan ciri. Akhir sekali, percubaan yang meluas dijalankan pada beberapa set data awam, dan keputusan menunjukkan bahawa DVLR yang dicadangkan kami mencapai prestasi yang lebih baik daripada beberapa kaedah pemilihan ciri tanpa pengawasan yang canggih.
Zihao SONG
Yantai University
Peng SONG
Yantai University
Chao SHENG
Yantai University
Wenming ZHENG
Southeast University
Wenjing ZHANG
Yantai University
Shaokai LI
Yantai University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Salinan
Zihao SONG, Peng SONG, Chao SHENG, Wenming ZHENG, Wenjing ZHANG, Shaokai LI, "A Novel Discriminative Virtual Label Regression Method for Unsupervised Feature Selection" in IEICE TRANSACTIONS on Information,
vol. E105-D, no. 1, pp. 175-179, January 2022, doi: 10.1587/transinf.2021EDL8067.
Abstract: Unsupervised Feature selection is an important dimensionality reduction technique to cope with high-dimensional data. It does not require prior label information, and has recently attracted much attention. However, it cannot fully utilize the discriminative information of samples, which may affect the feature selection performance. To tackle this problem, in this letter, we propose a novel discriminative virtual label regression method (DVLR) for unsupervised feature selection. In DVLR, we develop a virtual label regression function to guide the subspace learning based feature selection, which can select more discriminative features. Moreover, a linear discriminant analysis (LDA) term is used to make the model be more discriminative. To further make the model be more robust and select more representative features, we impose the ℓ2,1-norm on the regression and feature selection terms. Finally, extensive experiments are carried out on several public datasets, and the results demonstrate that our proposed DVLR achieves better performance than several state-of-the-art unsupervised feature selection methods.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2021EDL8067/_p
Salinan
@ARTICLE{e105-d_1_175,
author={Zihao SONG, Peng SONG, Chao SHENG, Wenming ZHENG, Wenjing ZHANG, Shaokai LI, },
journal={IEICE TRANSACTIONS on Information},
title={A Novel Discriminative Virtual Label Regression Method for Unsupervised Feature Selection},
year={2022},
volume={E105-D},
number={1},
pages={175-179},
abstract={Unsupervised Feature selection is an important dimensionality reduction technique to cope with high-dimensional data. It does not require prior label information, and has recently attracted much attention. However, it cannot fully utilize the discriminative information of samples, which may affect the feature selection performance. To tackle this problem, in this letter, we propose a novel discriminative virtual label regression method (DVLR) for unsupervised feature selection. In DVLR, we develop a virtual label regression function to guide the subspace learning based feature selection, which can select more discriminative features. Moreover, a linear discriminant analysis (LDA) term is used to make the model be more discriminative. To further make the model be more robust and select more representative features, we impose the ℓ2,1-norm on the regression and feature selection terms. Finally, extensive experiments are carried out on several public datasets, and the results demonstrate that our proposed DVLR achieves better performance than several state-of-the-art unsupervised feature selection methods.},
keywords={},
doi={10.1587/transinf.2021EDL8067},
ISSN={1745-1361},
month={January},}
Salinan
TY - JOUR
TI - A Novel Discriminative Virtual Label Regression Method for Unsupervised Feature Selection
T2 - IEICE TRANSACTIONS on Information
SP - 175
EP - 179
AU - Zihao SONG
AU - Peng SONG
AU - Chao SHENG
AU - Wenming ZHENG
AU - Wenjing ZHANG
AU - Shaokai LI
PY - 2022
DO - 10.1587/transinf.2021EDL8067
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E105-D
IS - 1
JA - IEICE TRANSACTIONS on Information
Y1 - January 2022
AB - Unsupervised Feature selection is an important dimensionality reduction technique to cope with high-dimensional data. It does not require prior label information, and has recently attracted much attention. However, it cannot fully utilize the discriminative information of samples, which may affect the feature selection performance. To tackle this problem, in this letter, we propose a novel discriminative virtual label regression method (DVLR) for unsupervised feature selection. In DVLR, we develop a virtual label regression function to guide the subspace learning based feature selection, which can select more discriminative features. Moreover, a linear discriminant analysis (LDA) term is used to make the model be more discriminative. To further make the model be more robust and select more representative features, we impose the ℓ2,1-norm on the regression and feature selection terms. Finally, extensive experiments are carried out on several public datasets, and the results demonstrate that our proposed DVLR achieves better performance than several state-of-the-art unsupervised feature selection methods.
ER -