A Novel Low-Rank Embedded Latent Multi-View Subspace Clustering Approach

Noises and outliers often degrade the final prediction performance in practical data processing. Multi-view learning by integrating complementary information across heterogeneous modalities has become one of the core techniques in the field of machine learning. However, existing methods rely on expl...

Full description

Saved in:
Bibliographic Details
Main Authors: Sen Wang, Lian Chen, Zhijian Liang, Qingyang Liu
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/9/2778
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850030718333222912
author Sen Wang
Lian Chen
Zhijian Liang
Qingyang Liu
author_facet Sen Wang
Lian Chen
Zhijian Liang
Qingyang Liu
author_sort Sen Wang
collection DOAJ
description Noises and outliers often degrade the final prediction performance in practical data processing. Multi-view learning by integrating complementary information across heterogeneous modalities has become one of the core techniques in the field of machine learning. However, existing methods rely on explicit-view clustering and stringent alignment assumptions, which affect the effectiveness in addressing the challenges such as inconsistencies between views, noise interference, and misalignment across different views. To alleviate these issues, we present a latent multi-view representation learning model based on low-rank embedding by implicitly uncovering the latent consistency structure of data, which allows us to achieve robust and efficient multi-view feature fusion. In particular, we utilize low-rank constraints to construct a unified latent subspace representation and introduce an adaptive noise suppression mechanism that significantly enhances robustness against outliers and noise interference. Moreover, the Augmented Lagrangian Multiplier Alternating Direction Minimization (ALM-ADM) framework enables efficient optimization of the proposed method. Experimental results on multiple benchmark datasets demonstrate that the proposed approach outperforms existing state-of-the-art methods in both clustering performance and robustness.
format Article
id doaj-art-4a2bd5b66a1a4d02a0509fb9c77d53de
institution DOAJ
issn 1424-8220
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-4a2bd5b66a1a4d02a0509fb9c77d53de2025-08-20T02:59:08ZengMDPI AGSensors1424-82202025-04-01259277810.3390/s25092778A Novel Low-Rank Embedded Latent Multi-View Subspace Clustering ApproachSen Wang0Lian Chen1Zhijian Liang2Qingyang Liu3School of Science, East China Jiaotong University, Nanchang 330013, ChinaSchool of Science, East China Jiaotong University, Nanchang 330013, ChinaSchool of Science, East China Jiaotong University, Nanchang 330013, ChinaSchool of Science, East China Jiaotong University, Nanchang 330013, ChinaNoises and outliers often degrade the final prediction performance in practical data processing. Multi-view learning by integrating complementary information across heterogeneous modalities has become one of the core techniques in the field of machine learning. However, existing methods rely on explicit-view clustering and stringent alignment assumptions, which affect the effectiveness in addressing the challenges such as inconsistencies between views, noise interference, and misalignment across different views. To alleviate these issues, we present a latent multi-view representation learning model based on low-rank embedding by implicitly uncovering the latent consistency structure of data, which allows us to achieve robust and efficient multi-view feature fusion. In particular, we utilize low-rank constraints to construct a unified latent subspace representation and introduce an adaptive noise suppression mechanism that significantly enhances robustness against outliers and noise interference. Moreover, the Augmented Lagrangian Multiplier Alternating Direction Minimization (ALM-ADM) framework enables efficient optimization of the proposed method. Experimental results on multiple benchmark datasets demonstrate that the proposed approach outperforms existing state-of-the-art methods in both clustering performance and robustness.https://www.mdpi.com/1424-8220/25/9/2778low-rank embeddingmulti-view learningheterogeneous feature fusionlatent space learning
spellingShingle Sen Wang
Lian Chen
Zhijian Liang
Qingyang Liu
A Novel Low-Rank Embedded Latent Multi-View Subspace Clustering Approach
Sensors
low-rank embedding
multi-view learning
heterogeneous feature fusion
latent space learning
title A Novel Low-Rank Embedded Latent Multi-View Subspace Clustering Approach
title_full A Novel Low-Rank Embedded Latent Multi-View Subspace Clustering Approach
title_fullStr A Novel Low-Rank Embedded Latent Multi-View Subspace Clustering Approach
title_full_unstemmed A Novel Low-Rank Embedded Latent Multi-View Subspace Clustering Approach
title_short A Novel Low-Rank Embedded Latent Multi-View Subspace Clustering Approach
title_sort novel low rank embedded latent multi view subspace clustering approach
topic low-rank embedding
multi-view learning
heterogeneous feature fusion
latent space learning
url https://www.mdpi.com/1424-8220/25/9/2778
work_keys_str_mv AT senwang anovellowrankembeddedlatentmultiviewsubspaceclusteringapproach
AT lianchen anovellowrankembeddedlatentmultiviewsubspaceclusteringapproach
AT zhijianliang anovellowrankembeddedlatentmultiviewsubspaceclusteringapproach
AT qingyangliu anovellowrankembeddedlatentmultiviewsubspaceclusteringapproach
AT senwang novellowrankembeddedlatentmultiviewsubspaceclusteringapproach
AT lianchen novellowrankembeddedlatentmultiviewsubspaceclusteringapproach
AT zhijianliang novellowrankembeddedlatentmultiviewsubspaceclusteringapproach
AT qingyangliu novellowrankembeddedlatentmultiviewsubspaceclusteringapproach