Article
Computer Science, Interdisciplinary Applications
Bastian Pfeifer, Marcus D. Bloice, Michael G. Schimek
Summary: Multi-view clustering methods are crucial for stratifying patients into sub-groups based on similar molecular characteristics. We introduce Parea, a multi-view hierarchical ensemble clustering approach that outperforms the current state-of-the-art on six out of seven analyzed cancer types. We have integrated the Parea method into our Python package Pyrea (https://github.com/mdbloice/Pyrea), which facilitates the effortless and flexible design of ensemble workflows incorporating various fusion and clustering algorithms.
JOURNAL OF BIOMEDICAL INFORMATICS
(2023)
Article
Mathematics
Jianwen Gan, Yunhui Liang, Liang Du
Summary: The clustering ensemble method has gained attention for its ability to enhance the stability and robustness of single clustering methods. Among these methods, similarity-matrix-based or graph-based approaches have been widely applied. However, most similarity-matrix-based methods overlook the importance of relevance ranking within the same cluster, leading to unreliable similarity estimates and degraded clustering performance. In this paper, we propose a graph-learning-based ensemble algorithm that optimizes adaptive weights based on the ranking importance of different neighbors, resulting in improved accuracy compared to related clustering methods.
Review
Computer Science, Information Systems
Maryam Abdolali, Nicolas Gillis
Summary: Subspace clustering is an important unsupervised clustering approach that assumes high-dimensional data points are distributed around low-dimensional linear subspaces. To overcome linearity restrictions, various nonlinear approaches have been proposed in the last decade. New taxonomy classifies approaches into locality preserving, kernel based, and neural network based categories, with detailed comparisons on synthetic and real-world data sets revealing potential directions for future research and unresolved challenges.
COMPUTER SCIENCE REVIEW
(2021)
Article
Automation & Control Systems
Guoqing Liu, Hongwei Ge, Ting Li, Shuzhi Su, Shuangxi Wang
Summary: We propose a robust multi-view subspace enhancement representation algorithm, named MSC, based on collaborative constraints and Hilbert-Schmidt independence criterion (HSIC) induction method, to improve the recognition performance and anti-noise interference ability. HSIC is used as a diversity regularization term to mine the complementary information between different views. Sparse constraint, hypergraph regularization, and low-rank idea are introduced to enhance the diagonal block structure and capture the local geometric structure. Experimental results demonstrate that the proposed method achieves competitive performance compared to state-of-the-art multi-view clustering methods, with and without noisy views.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Information Systems
Siyuan Zhao
Summary: Sparse subspace clustering is a widely used method for clustering high dimensional data. However, the traditional method is complex and requires prior information. In this paper, we propose a new method called Self-constrained Sparse Subspace Clustering (ScSSC) to simplify the clustering of high dimensional data. The proposed algorithm is a non-deep neural network model that can discover a high-quality cluster structure without prior information, making it highly effective in unsupervised scenarios.
Article
Computer Science, Information Systems
Na Yu, Yusen Zhang, Rui Gao
Summary: This paper proposes a novel tensor method, called enhanced tensor nuclear norm and hypergraph Laplacian regularization (ETHLR), to address the degradation of conventional tensor robust principal component analysis (TRPCA). ETHLR can jointly learn the prior knowledge of singular values and high-order manifold structures. Enhanced tensor nuclear norm is used to shrink the singular values and consider their difference information, while hypergraph Laplacian constraint helps encode high-order geometric structures. Experimental results on pan-cancer omics data demonstrate the superiority of ETHLR over several state-of-the-art competitors.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Computer Science, Artificial Intelligence
Libin Wang, Yulong Wang, Hao Deng, Hong Chen
Summary: This paper proposes a robust sparse subspace clustering method called ARSSC, which assigns small weights to corrupted entries in each data point, reducing the attention to them. Non-convex penalties are also utilized to overcome the overpenalized problem. The effectiveness of the proposed method is validated through experiments on real-world databases.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Artificial Intelligence
Qiaoping Wang, Xiaoyun Chen, Wenjian Chen
Summary: In this paper, a novel multi-view subspace clustering method called ARMSC is proposed, which utilizes auto-weighted graph regularization and residual compensation to improve clustering performance. Experimental results demonstrate the superiority of ARMSC over several state-of-the-art multi-view clustering approaches on various real-world datasets.
NEURAL PROCESSING LETTERS
(2022)
Article
Computer Science, Information Systems
Yingxu Wang, Long Chen, Jin Zhou, Tianjun Li, Yufeng Yu
Summary: This paper proposes a new pairwise constraints-based semi-supervised fuzzy clustering method with multi-manifold regularization (MMRFCM), which can overcome the deficiencies of current methods and achieve excellent clustering results.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Qingjiang Xiao, Shiqiang Du, Yao Yu, Yixuan Huang, Jinmei Song
Summary: The proposed JWHMSC model integrates representation learning, weighted tensor nuclear norm (WTNN) constraint, and hyper-Laplacian graph regularization constraint to accurately obtain sample distribution of classification information in multi-view subspace clustering. Experimental results on eight benchmark datasets show the good performance of JWHMSC compared to state-of-the-art methods.
JOURNAL OF INTELLIGENT & FUZZY SYSTEMS
(2022)
Article
Biochemical Research Methods
Juan Wang, Cong-Hai Lu, Xiang-Zhen Kong, Ling-Yun Dai, Shasha Yuan, Xiaofeng Zhang
Summary: The identification of cancer types is crucial for early diagnosis and treatment of cancer. This study proposes a new low-rank subspace clustering method (MmCLRR) to effectively cluster cancer types by utilizing complementary information from cancer multi-omics data. Experimental results demonstrate its superiority in multi-view clustering.
BMC BIOINFORMATICS
(2022)
Article
Biochemical Research Methods
Jing Zhao, Bowen Zhao, Xiaotong Song, Chujun Lyu, Weizhi Chen, Yi Xiong, Dong-Qing Wei
Summary: The Subtype-DCC method, which integrates multi-omics data, is proposed for cancer subtyping and demonstrates superior performance compared to existing clustering methods. It has potential applications in cancer diagnosis, prognosis, and treatment.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Computer Science, Interdisciplinary Applications
Bhavana Bansal, Anita Sahoo
Summary: Cancer subtype discovery is crucial for personalized treatment. This paper proposes an improved sparse-jNMF framework combined with consensus clustering for predicting homogeneous subgroups of cancer patients.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2023)
Article
Genetics & Heredity
Jie Feng, Limin Jiang, Shuhao Li, Jijun Tang, Lan Wen
Summary: The multiple sources of cancer determine its multiple causes and various subtypes, which are key in personalized cancer treatment and clinical diagnosis. Through gene expression profiling and clustering algorithms, different cancer subtypes can be classified and studied effectively.
FRONTIERS IN GENETICS
(2021)
Article
Engineering, Electrical & Electronic
Yin-Ping Zhao, Long Chen, C. L. Philip Chen
Summary: This paper proposes a novel model called Laplacian regularized nonnegative representation (LapNR), which ensures that the query sample should be approximated from homogeneous samples and irrelevant to heterogeneous ones. By applying the graph Laplacian to the nonnegative representations, the model captures geometric information of input data and generates a sparse and discriminative representation matrix. The efficient optimization procedure based on the alternating direction method of multipliers (ADMM) is developed for LapNR and the extensive experiments demonstrate its effectiveness and efficiency in clustering and dimensionality reduction tasks.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2021)
Article
Biochemical Research Methods
Weiwen Wang, Xiwen Zhang, Dao-Qing Dai
Summary: This study introduces a multi-omics integration method that considers noise effects and data-specific patterns, utilizing denoised network regularization to capture data-specific patterns, which outperforms other methods in discovering common patterns among data types.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Weiwen Wang, Xiwen Zhang, Dao-Qing Dai
Summary: This study presents a method called springD(2)A for drug repositioning, which can capture the uncertainty in negative pairs and discriminate between positive and unknown pairs, improving the reliability of predictions.
Article
Biochemical Research Methods
Xiwen Zhang, Weiwen Wang, Chuan-Xian Ren, Dao-Qing Dai
Summary: This article introduces a representation learning method for multiple biological networks. The method utilizes denoised diffusion and graph regularized integration to handle noise and spurious edges, while preserving the common structure and specific information of different networks, resulting in useful representation features.
BRIEFINGS IN BIOINFORMATICS
(2022)
Proceedings Paper
Acoustics
Yuqing Liu, Weiwen Wang, Chuan-Xian Ren, Dao-Qing Dai
Summary: The achievements of deep learning in image processing provide the possibility of using pathological images for MSI detection, but traditional deep networks cannot achieve satisfactory performance and generalize well, hence the proposed MetaCon model which shows superiority in experiments on two public datasets.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT VIII
(2021)