Article
Health Care Sciences & Services
Jihwan Ha
Summary: This article presents a novel computational framework (MDMF) for identifying potential miRNA-disease associations. By employing matrix factorization with a disease similarity constraint, MDMF efficiently discovers potential miRNA-disease associations and deciphers the underlying roles of miRNAs in disease pathogenesis at a system level.
JOURNAL OF PERSONALIZED MEDICINE
(2022)
Article
Biochemistry & Molecular Biology
Linqian Cui, You Lu, Jiacheng Sun, Qiming Fu, Xiao Xu, Hongjie Wu, Jianping Chen
Summary: Multiple studies have demonstrated the crucial role of microRNAs in the research of complex human diseases, and the limitations of traditional biological experiments have highlighted the need for computational simulation to predict unknown miRNA-disease associations. By combining reinforcement learning's Q-learning algorithm, a RFLMDA model was proposed to obtain an optimal weight S by fusing three submodels. Experimental results indicated that the RFLMDA model outperformed other methods in predicting miRNA-disease relationships.
Article
Computer Science, Artificial Intelligence
Jihwan Ha
Summary: A simple and effective computational framework, SMAP, was proposed to identify miRNA-disease associations by applying the recommended algorithm with miRNA and disease similarity constraints. SMAP utilized known miRNA-disease associations and integrated similarities for miRNAs and diseases. The AUCs of SMAP in global and local leave-one-out cross validation were 0.9227 and 0.8952, respectively. Independent case studies on two major human cancers confirmed the comparable performance of SMAP. Overall, SMAP could serve as an effective guide to understanding the pathogenesis and etiology of human diseases and deciphering potential disease biomarkers.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Genetics & Heredity
Jiancheng Ni, Lei Li, Yutian Wang, Cunmei Ji, Chunhou Zheng
Summary: This study presents a new algorithm, MDSCMF, to predict the associations between miRNAs and diseases. The algorithm utilizes matrix decomposition and similarity-constrained matrix factorization to integrate different similarity measures and determine the similarity between miRNAs and diseases. Experimental results demonstrate the effectiveness of MDSCMF in predicting miRNA-disease associations.
Article
Computer Science, Artificial Intelligence
Jovan Chavoshinejad, Seyed Amjad Seyedi, Fardin Akhlaghian Tab, Navid Salahian
Summary: Semi-supervised nonnegative matrix factorization combines the strengths of matrix factorization in learning part-based representation and can achieve high learning performance with limited labeled data and a large amount of unlabeled data. Recent research focuses on utilizing self-supervised learning to enhance semi-supervised learning. This paper proposes an effective Self-Supervised Semi-Supervised Nonnegative Matrix Factorization (S4NMF) model that directly extracts a consensus result from ensembled NMFs with similarity and dissimilarity regularizations. Experimental results on standard benchmark datasets demonstrate the effectiveness of the proposed model in semi-supervised clustering.
PATTERN RECOGNITION
(2023)
Article
Automation & Control Systems
Zhiwei Xing, Meng Wen, Jigen Peng, Jinqian Feng
Summary: The paper introduces a novel discriminative semi-supervised NMF (DSSNMF) algorithm that effectively utilizes label information from a portion of the data, with empirical experiments demonstrating its effectiveness.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2021)
Article
Computer Science, Artificial Intelligence
Wenhui Wu, Junhui Hou, Shiqi Wang, Sam Kwong, Yu Zhou
Summary: In this paper, we propose a semi-supervised adaptive kernel concept factorization (SAKCF) method that integrates data representation and kernel learning and solves the problem using an alternating iterative algorithm. Experimental results demonstrate the effectiveness and advantages of SAKCF over other methods in clustering tasks.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Artificial Intelligence
Jingxing Yin, Siyuan Peng, Zhijing Yang, Badong Chen, Zhiping Lin
Summary: A new semi-supervised symmetric nonnegative matrix factorization (SNMF) method, called hypergraph based semi-supervised SNMF (HSSNMF), is proposed for image clustering. HSSNMF constructs a similarity matrix using a predefined hypergraph and propagates pairwise constraints using a hypergraph-based algorithm. The discriminative assignment matrix is obtained through optimization. Experimental results demonstrate the superiority of HSSNMF compared to other state-of-the-art methods.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Information Systems
Fan Yang, Xiaojian Ding, Fumin Ma, Deyu Tong, Jie Cao
Summary: Hashing has attracted much attention in multi-modal similarity search applications, but most existing approaches suffer from large quantization errors due to their reliance on relaxation schemes to generate binary codes. In addition, embedding labels into pairwise similarity matrix in existing approaches leads to expensive time and space costs and the loss of category information. To address these issues, we propose Efficient Discrete Matrix factorization Hashing (EDMH). Our approach learns latent subspaces for individual modality through matrix factorization, preserving semantic structure representation information. We also introduce an efficient discrete optimization scheme to generate compact binary codes. Experimental results demonstrate that our proposed algorithms outperform state-of-the-art approaches, achieving average improvements of 2.50% (for Wiki), 2.66% (for MIRFlickr), and 2.25% (for NUS-WIDE).
INFORMATION PROCESSING & MANAGEMENT
(2023)
Review
Computer Science, Artificial Intelligence
Wen-Sheng Chen, Kexin Xie, Rui Liu, Binbin Pan
Summary: This paper focuses on the theoretical idea, basic model, optimization method, and variants of symmetric non-negative matrix factorization (SNMF), a promising tool for data analysis. It classifies SNMF-related approaches into classic SNMFs and extended SNMFs, elaborates on key concepts, characteristics, and current issues of these algorithms, and compares the clustering performance and algorithm effects of SNMF and its variants on object image datasets. Additionally, it compares the performance of similarity matrix construction methods and discusses open problems with SNMF.
Article
Computer Science, Artificial Intelligence
Zhiwei Xing, Yingcang Ma, Xiaofei Yang, Feiping Nie
Summary: This paper presents a new nonnegative matrix factorization method, GNMFLD, under graph and label constraints to enhance clustering discrimination and performance. Empirical experiments demonstrate the effectiveness of the proposed method.
Article
Computer Science, Information Systems
Sheng Bi, Xiangli Li
Summary: This paper presents a semi-supervised clustering ensemble based on the genetic algorithm model, which enhances the algorithm's performance by utilizing pairwise constraint information.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Siyuan Peng, Zhijing Yang, Bingo Wing-Kuen Ling, Badong Chen, Zhiping Lin
Summary: A new semi-supervised NMF method called dual semi-supervised convex nonnegative matrix factorization (DCNMF) is proposed in this paper. DCNMF incorporates the pointwise and pairwise constraints of labeled samples into convex NMF, resulting in a better low-dimensional data representation. It can process mixed-sign data due to the nonnegative constraint only on the coefficient matrix.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Ying Zhang, Xiangli Li, Mengxue Jia
Summary: Traditional clustering is an unsupervised learning method, but prior information in actual data can be used for semi-supervised clustering. Pairwise constraints are commonly used prior information that can improve clustering performance. This paper proposes a semi-supervised clustering method that combines pairwise constraints with nonnegative matrix factorization and verifies its effectiveness through experiments.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2022)
Article
Automation & Control Systems
Kexin Zhang, Xuezhuan Zhao, Siyuan Peng
Summary: A novel multiple graph regularized semi-supervised NMF method, MSNMF, is proposed in this paper, which combines limited supervised information in the form of pairwise constraints with multiple graph regularization to capture discriminative data representation. Experimental results on eight practical image datasets demonstrate that MSNMF can achieve better clustering results than several related NMF methods.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2021)
Article
Biochemical Research Methods
Chengqian Lu, Lishen Zhang, Min Zeng, Wei Lan, Guihua Duan, Jianxin Wang
Summary: Emerging evidence suggests that circRNAs, with their covalently closed loop structures, can serve as promising biomarkers for diagnosis in pathogenic processes. Computational approaches provide a cost-effective way to identify circRNA-disease associations by aggregating multi-source pathogenesis data and inferring potential associations at the system level. The proposed CDHGNN model, based on edge-weighted graph attention and heterogeneous graph neural networks, outperforms state-of-the-art algorithms in predicting circRNA-disease associations and can identify specific molecular associations and investigate biomolecular regulatory relationships in pathogenesis.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biochemical Research Methods
Lishen Zhang, Chengqian Lu, Min Zeng, Yaohang Li, Jianxin Wang
Summary: In this study, we propose a method called CRMSS for discriminating circRNA-RBP binding sites based on multi-scale characterizing sequence and structure features. The CRMSS achieves superior performance over state-of-the-art methods in predicting circRNA-RBP binding.
BRIEFINGS IN BIOINFORMATICS
(2023)
Review
Biochemical Research Methods
Wei Lan, Yi Dong, Hongyu Zhang, Chunling Li, Qingfeng Chen, Jin Liu, Jianxin Wang, Yi-Ping Phoebe Chen
Summary: Accumulating evidence shows the importance of circular RNA (circRNA) in human diseases. Computational methods have been proposed to identify circRNA-disease associations, but there is a lack of comprehensive comparisons and summaries of these methods. This paper categorizes existing methods into three groups and introduces baseline methods for each category. It compares 14 representative methods using 5 different datasets and evaluates their effectiveness in identifying circRNA-disease associations in common cancers. The study also discusses the observations about method robustness and future directions and challenges.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biochemical Research Methods
Yiran Huang, Yongjin Bin, Pingfan Zeng, Wei Lan, Cheng Zhong
Summary: Drug repositioning is an important approach for predicting new disease indications for existing drugs. This paper proposes a neighborhood interaction-based method called NetPro for drug repositioning via label propagation. Experimental results show that NetPro can effectively identify potential drug-disease associations and achieve better prediction performance than existing methods. Case studies demonstrate that NetPro is capable of predicting promising candidate disease indications for drugs.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)
Article
Computer Science, Hardware & Architecture
Jingling Liu, Jiawei Huang, Weihe Li, Jianxin Wang, Tian He
Summary: Datacenter networks often face path asymmetry, leading to problems like packet reordering and under-utilization of multiple paths. In this paper, we propose a load balancing mechanism called AG that adaptively adjusts switching granularity based on the degree of topology asymmetry. We also design a switch-based scheme to measure the difference of one-way delay, allowing accurate detection of topology asymmetry. Experimental results show that AG outperforms existing load balancing schemes in terms of flow completion time.
IEEE-ACM TRANSACTIONS ON NETWORKING
(2023)
Article
Biochemical Research Methods
Yanping Zeng, Yongxin He, Ruiqing Zheng, Min Li
Summary: Gene regulatory network plays a crucial role in controlling biological processes. Deciphering complex gene regulatory networks remains challenging. Recent advances in single-cell RNA sequencing enable computational inference of cell-specific gene regulatory networks. Normi is a novel gene regulatory network inference method that addresses challenges of pseudo-time information and dropout data. Normi outperforms other methods and identifies key regulators and crucial biological processes.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biochemical Research Methods
Xuhua Yan, Ruiqing Zheng, Fangxiang Wu, Min Li
Summary: CIAIRE is a novel contrastive learning-based batch correction framework that achieves a superior mix-heterogeneity trade-off. It proposes two complementary strategies, construction strategy and refinement strategy, to improve the appropriateness of positive pairs. CLAIRE outperforms existing methods in terms of mix-heterogeneity trade-off and achieves the best integration performance on six real datasets.
Article
Computer Science, Information Systems
Weihe Li, Jiawei Huang, Wenjun Lyu, Baoshen Guo, Wanchun Jiang, Jianxin Wang
Summary: Current ABR algorithms do not pay enough attention to audio bitrate selection, assuming it has minimal impact on video selection. However, with the advancement of audio technologies, audio bitrate can significantly affect video selection and viewing experience. To address this issue, we propose a deep reinforcement learning-based ABR algorithm that considers both audio and video quality, achieving significant improvements in overall viewing quality.
IEEE TRANSACTIONS ON MULTIMEDIA
(2023)
Article
Computer Science, Information Systems
Shigeng Zhang, Zijing Ma, Kaixuan Lu, Xuan Liu, Jia Liu, Song Guo, Albert Y. Zomaya, Jian Zhang, Jianxin Wang
Summary: This paper presents HearMe, an accurate and real-time lip-reading system built on commercial RFID devices. HearMe can help people with speech disorders communicate and interact with the world effectively. By utilizing effective data collection, signal pattern extraction, and feature extraction techniques, HearMe achieves high accuracy in mouth motion recognition and word-level recognition. The use of transfer learning enhances the model's robustness in different environments.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2023)
Article
Computer Science, Information Systems
Shigeng Zhang, Zijing Ma, Chengwei Yang, Xiaoyan Kui, Xuan Liu, Weiping Wang, Jianxin Wang, Song Guo
Summary: This article introduces a real-time and accurate gesture recognition system called ReActor based on RFID. ReActor combines time-domain statistical features and frequency-domain features to represent the signal profile corresponding to different gestures accurately, and uses signal preprocessing and classifier training to maintain high accuracy in different environments.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2023)
Article
Multidisciplinary Sciences
Huimin Zhu, Renyi Zhou, Dongsheng Cao, Jing Tang, Min Li
Summary: This article introduces a deep learning approach called PGMG, guided by pharmacophore, to generate bioactive molecules. The molecules generated by PGMG have strong docking affinities and high scores of validity, uniqueness, and novelty.
NATURE COMMUNICATIONS
(2023)
Article
Computer Science, Theory & Methods
Wanchun Jiang, Haoyang Li, Yulong Yan, Fa Ji, Jiawei Huang, Jianxin Wang, Tong Zhang
Summary: This paper investigates the scheduling problem of key-value access operations in distributed key-value stores and proposes a distributed adaptive scheduler (DAS). Theoretical analysis and extensive simulations show that DAS can adapt to varying traffic and server performance, achieving consistent low latency and shorter request completion time.
IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS
(2023)
Article
Biochemical Research Methods
Yiming Li, Min Zeng, Fuhao Zhang, Fang-Xiang Wu, Min Li
Summary: In this study, DeepCellEss, a sequence-based interpretable deep learning framework, is proposed for cell line-specific essential protein predictions. By utilizing convolutional neural networks, bidirectional long short-term memory, and multi-head self-attention mechanism, DeepCellEss achieves effective prediction performance for different cell lines and outperforms existing methods and metrics.