Article
Computer Science, Artificial Intelligence
Dimosthenis Pasadakis, Christie Louis Alappat, Olaf Schenk, Gerhard Wellein
Summary: The study introduces a novel direct multiway spectral clustering algorithm based on p-norm, using a nonlinear reformulation of the spectral clustering method to achieve improved numerical benefits within a certain range. By recasting the problem and promoting sparser solution vectors, it aims to achieve optimal graph cuts as p approaches one.
Article
Automation & Control Systems
Nicolas Garcia Trillos, Pengfei He, Chenghui Li
Summary: In this work, statistical properties of graph-based algorithms for multi-manifold clustering are studied, and sufficient conditions for similarity graphs to capture the right geometric information to solve the problem are provided. An example and extensive numerical experiments provide insights on the multi-manifold clustering problem.
JOURNAL OF MACHINE LEARNING RESEARCH
(2023)
Article
Computer Science, Artificial Intelligence
Joel Mackenzie, Matthias Petri, Alistair Moffat
Summary: Connectivity graphs and linked data are crucial for the information access systems in social media and web search services. The Bipartite Graph Partitioning (BP) mechanism enhances compressibility and reduces storage space occupied by large sparse graphs. We have developed algorithmic and heuristic refinements to BP, resulting in faster computation of space-reducing vertex orderings. Our implementation executes up to four times faster than the baseline implementation, while maintaining compressibility.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Information Systems
Bo Zhou, Wenliang Liu, Wenzhen Zhang, Zhengyu Lu, Qianlin Tan
Summary: In this paper, a new method of multi-kernel graph fusion (MKGF-MM) based on min-max optimization is proposed for spectral clustering, which fully utilizes all base kernels to address the bias to power base kernels issue. The proposed method investigates a novel min-max weight strategy to capture the complementary information among all base kernels and designs an iterative optimization method to solve the objective function. The theoretical proof of convergence is provided and experimental results demonstrate the superiority of the proposed method over comparison methods, along with fast convergence of the proposed optimization method.
INFORMATION PROCESSING & MANAGEMENT
(2022)
Article
Computer Science, Artificial Intelligence
Ling Ding, Shifei Ding, Yanru Wang, Lijuan Wang, Hongjie Jia
Summary: The study proposes a new manifold p-spectral clustering algorithm (M-pSC) using path-based affinity measure to handle manifold data, constructing more accurate affinity matrix and improving clustering quality and robustness.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2021)
Article
Computer Science, Artificial Intelligence
Qilin Li, Senjian An, Ling Li, Wanquan Liu, Yanda Shao
Summary: This paper presents a novel approach to multi-view graph learning by combining weight learning and graph learning in an alternating optimization framework. The approach fuses multiple affinity graphs using a weight learning scheme based on unsupervised graph smoothness and utilizes them as a consensus prior to diffusion. It introduces a multi-view diffusion process that learns a manifold-aware affinity graph by propagating affinities on tensor product graphs, improving pairwise affinities using higher-order contextual information. The proposed approach outperforms state-of-the-art methods for image retrieval and clustering on 13 out of 16 real-world datasets.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2023)
Article
Computer Science, Artificial Intelligence
Guo Zhong, Ting Shu, Guoheng Huang, Xueming Yan
Summary: Multi-view spectral clustering methods often require multiple separate steps, leading to compromised clustering performance. This work proposes a method that simultaneously performs consensus graph learning and discretization, improving clustering performance by avoiding information loss among independent steps.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Information Systems
Suhang Gu, Fu-Lai Chung, Shitong Wang
Summary: The novel algorithm MVCPMM aims to achieve more consistent multi-view clustering results with only one random initialization and one parameter. By adding virtual function nodes to pass smooth messages between different views, MVCPMM improves clustering quality and consistency. Experimental results demonstrate the superiority of MVCPMM in terms of clustering performance and consistency across different views.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Jingwei Chen, Jianyong Zhu, Shiyu Xie, Hui Yang, Feiping Nie
Summary: A new framework is proposed in this study to jointly perform spectral embedding and improve spectral rotation using an anchor-based acceleration strategy. Experimental results validate the effectiveness of the proposed algorithm across multiple datasets.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Xiaotong Zhang, Han Liu, Xiao-Ming Wu, Xianchao Zhang, Xinyue Liu
Summary: This paper proposes a spectral embedding network named SENet for attributed graph clustering, which improves graph structure and learns node embeddings to address issues in traditional methods, achieving superior performance over state-of-the-art methods.
Article
Computer Science, Information Systems
Jingzhi Tu, Gang Mei, Francesco Piccialli
Summary: This paper proposes an improved Nystrom spectral graph clustering method based on k-core decomposition sampling for large networks. The method utilizes k-core decomposition as a sampling method, conducts NSC on the samples to acquire clusters, and then applies a label propagation algorithm to group the remaining nodes. Experimental results show that the proposed method achieves higher accuracy than NSC and higher efficiency than spectral clustering.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Guo Zhong, Chi-Man Pun
Summary: By integrating multiple views, multi-view learning can improve the performance of learning tasks by discovering underlying data structures. Multi-view clustering, as a basic and important branch of multi-view learning, has achieved great success recently. However, most existing multi-view spectral clustering methods may significantly affect clustering performance due to information loss between independent steps. In this paper, a novel Self-taught Multi-view Spectral Clustering (SMSC) framework is proposed to address this issue, and experimental results show that our methods out-perform other state-of-the-art baselines.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Artificial Intelligence
Zhen Wang, Zhaoqing Li, Rong Wang, Feiping Nie, Xuelong Li
Summary: This paper proposes a new method for graph clustering to solve challenging problems in spectral clustering methods by simultaneously performing spectral embedding and spectral rotation. By deriving a low-dimensional representation matrix from a graph using label propagation, a double-stochastic and positive semidefinite similarity matrix can be reconstructed, accelerating the algorithm. Experimental results demonstrate excellent performance of the method in terms of time cost and accuracy.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2021)
Article
Computer Science, Artificial Intelligence
Yang Zhang, Moyun Liu, Huiming Zhang, Guodong Sun, Jingwu He
Summary: Affinity graph-based segmentation methods have become a major trend in computer vision. This paper proposes an adaptive fusion affinity graph with noise-free low-rank representation in an online manner for natural image segmentation, which effectively removes noisy data and achieves excellent performance.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Information Systems
Abdul Atif Khan, Sraban Kumar Mohanty
Summary: Spectral clustering is a popular unsupervised learning technique used for exploratory analysis of complex datasets. In this paper, a parameter-free method is proposed to construct a sparse proximity graph, aiming to improve the computational efficiency of spectral clustering. Experimental results show that the proposed algorithm outperforms other spectral clustering methods in terms of cluster quality, computational time, and sparsity of the graph.
INFORMATION SCIENCES
(2022)
Article
Environmental Sciences
Na Li, Deyun Zhou, Jiao Shi, Mingyang Zhang, Tao Wu, Maoguo Gong
Summary: Due to the superior spatial-spectral extraction capability of CNN, it has great potential in DR of HSIs. However, most CNN-based methods are supervised, while unsupervised methods often focus on data reconstruction rather than discriminability. Therefore, a deep fully convolutional embedding network (DFCEN) is proposed in this study to address these issues and improve the effectiveness of unsupervised learning.
Article
Neurosciences
Matthew Brendel, Chang Su, Yu Hou, Claire Henchcliffe, Fei Wang
Summary: This study aimed to identify subtypes of moderate-to-advanced Parkinson's disease by comprehensively considering motor and non-motor manifestations. Three unique subtypes emerged from the clustering results, each characterized by different levels of symptom severity. These subtypes showed significant differences in motor and non-motor clinical features, providing important information for further research on the treatment and management of Parkinson's disease.
NPJ PARKINSONS DISEASE
(2021)
Article
Health Care Sciences & Services
Chang Su, Yongkang Zhang, James H. Flory, Mark G. Weiner, Rainu Kaushal, Edward J. Schenck, Fei Wang
Summary: The study identified 4 biologically distinct subphenotypes of COVID-19 using machine learning and clinical data, which were highly predictive of clinical outcomes. It found varying prevalence of subphenotypes across the peak of the outbreak in NYC. Furthermore, social determinants of health specifically influenced mortality outcomes in certain subphenotypes.
NPJ DIGITAL MEDICINE
(2021)
Article
Clinical Neurology
Ivan Guan, Maissa Trabilsy, Samantha Barkan, Ashwin Malhotra, Yu Hou, Fei Wang, Natalie Hellmers, Harini Sarva, Claire Henchcliffe
Summary: This study compared levodopa off/on testing with Parkinson's Kinetigraph motor scores in PD patients. The results showed that a robust off/on response does not necessarily indicate adequately controlled motor symptoms. The PKG may provide additional clinically relevant data on motor symptoms for prospective observational studies.
CLINICAL NEUROLOGY AND NEUROSURGERY
(2021)
Article
Computer Science, Interdisciplinary Applications
Zheng Yuan, Zhengyun Zhao, Haixia Sun, Jiao Li, Fei Wang, Sheng Yu
Summary: This paper introduces knowledge-aware embedding, CODER, a critical tool for medical term normalization. By utilizing contrastive learning and a medical knowledge graph, CODER can extract semantic similarity and relatedness of medical concepts, which can be used for medical term normalization or feature extraction for machine learning.
JOURNAL OF BIOMEDICAL INFORMATICS
(2022)
Article
Oncology
Zhaoyi Chen, Hansi Zhang, Thomas J. George, Yi Guo, Mattia Prosperi, Jingchuan Guo, Dejana Braithwaite, Fei Wang, Warren Kibbe, Lynne Wagner, Jiang Bian
Summary: In this study, we simulated colorectal cancer trials using real-world data and tested two simulation scenarios. The results showed that our simulations can generate effectiveness and safety outcomes comparable with the original trials.
JCO CLINICAL CANCER INFORMATICS
(2022)
Article
Environmental Sciences
Hongyu Zhao, Kaiyuan Feng, Yue Wu, Maoguo Gong
Summary: This paper proposes a novel feature extraction network that combines Recurrent Neural Network (RNN) and Convolutional Neural Network (CNN) for hyperspectral change detection tasks. The experimental results demonstrate that the proposed method yields reliable detection results and has fewer noise regions.
Article
Environmental Sciences
Junyu Gao, Maoguo Gong, Xuelong Li
Summary: In this paper, we propose a method named SwinCounter for object counting in remote sensing. The method addresses the issue of imbalanced object labels by introducing a Balanced MSE Loss and captures multi-scale information accurately using the attention mechanism. Experiments on the RSOC dataset demonstrate the competitiveness and superiority of the proposed method.
Review
Medicine, General & Internal
Jie Xu, Yunyu Xiao, Wendy Hui Wang, Yue Ning, Elizabeth A. Shenkman, Jiang Bian, Fei Wang
Summary: Machine learning models are increasingly being used in clinical decision-making, but recent research has highlighted the potential biases that these techniques may introduce, particularly for vulnerable ethnic minorities. This paper provides a comprehensive review of algorithmic fairness in computational medicine, discussing different types of bias, metrics for quantifying fairness, and methods for mitigating bias. It also summarizes popular software libraries and tools for evaluating and mitigating bias, serving as a valuable resource for researchers and practitioners in computational medicine.
Article
Computer Science, Information Systems
Zijun Yao, Bin Liu, Fei Wang, Daby Sow, Ying Li
Summary: This article proposes a novel prescription recommendation framework called OntoPath, which predicts the next drug in chronic disease treatment pathways by integrating multiple medical evidence from domain knowledge guidance, medical history profiling, and side information utilization. Extensive experiments on a large-scale depression cohort demonstrate the effectiveness of OntoPath in prescription recommendation.
ACM TRANSACTIONS ON INFORMATION SYSTEMS
(2023)
Review
Genetics & Heredity
Matthew Brendel, Chang Su, Zilong Bai, Hao Zhang, Olivier Elemento, Fei Wang
Summary: Single-cell RNA sequencing (scRNA-seq) is widely used to quantify the gene expression profile of thousands of single cells simultaneously. Deep learning techniques have emerged as a promising tool for scRNA-seq data analysis, allowing for the extraction of informative and compact features from noisy and high-dimensional data. This review surveys recent developments in deep learning for scRNA-seq data analysis, highlights key advancements made by deep learning in the analysis pipeline, and discusses the benefits and challenges of applying deep learning to scRNA-seq data.
GENOMICS PROTEOMICS & BIOINFORMATICS
(2022)
Review
Health Care Sciences & Services
Adrienne Kline, Hanyin Wang, Yikuan Li, Saya Dennis, Meghan Hutch, Zhenxing Xu, Fei Wang, Feixiong Cheng, Yuan Luo
Summary: This review summarizes current studies on multi-modal data fusion in the health sector, highlighting the common use of multi-modal methods in neurology and oncology and the improved predictive performance achieved through data fusion. However, the lack of clear clinical deployment strategies, FDA approval, and analysis of biases and healthcare disparities in diverse sub-populations was noted.
NPJ DIGITAL MEDICINE
(2022)
Article
Computer Science, Artificial Intelligence
Feng-Lei Fan, Mengzhou Li, Fei Wang, Rongjie Lai, Ge Wang
Summary: Inspired by the diversity of biological neurons, this paper explores the application of quadratic artificial neurons in deep learning and highlights the differences in expressivity and training risk between traditional neurons and quadratic networks with or without quadratic activation. By applying spline theory and algebraic geometry, the superior model expressivity of quadratic networks over traditional networks is mathematically demonstrated, and an effective training strategy called ReLinear is proposed to stabilize the training process of quadratic networks.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Proceedings Paper
Computer Science, Information Systems
Qian Yang, Yuexing Hao, Kexin Quan, Stephen Yang, Yiran Zhao, Volodymyr Kuleshov, Fei Wang
Summary: Clinical decision support tools (DSTs) powered by AI can improve diagnostic and treatment decision-making, but AI models are not always correct. To address this, researchers investigated how clinicians validate each other's suggestions and designed a new DST that incorporates these interactions. The design uses GPT-3 to provide literature evidence showcasing the robustness and applicability of AI suggestions. A prototype study with clinicians demonstrated the promise of this approach and revealed new opportunities for design and research.
PROCEEDINGS OF THE 2023 CHI CONFERENCE ON HUMAN FACTORS IN COMPUTING SYSTEMS (CHI 2023)
(2023)
Article
Health Care Sciences & Services
Zhaoyi Chen, Hansi Zhang, Yi Guo, Thomas J. George, Mattia Prosperi, William R. Hogan, Zhe He, Elizabeth A. Shenkman, Fei Wang, Jiang Bian
Summary: This study explored the feasibility of using real-world data to simulate clinical trials for Alzheimer's disease, comparing different formulations of donepezil. Two main simulation scenarios were considered: one-arm simulation and two-arm simulation with propensity score matching. Higher SAE rates were observed in the simulated trials compared to the original trial, indicating potential limitations of the approach.
NPJ DIGITAL MEDICINE
(2021)
Article
Computer Science, Artificial Intelligence
Guang-Hai Liu, Zuo-Yong Li, Jing-Yu Yang, David Zhang
Summary: This article introduces a novel image retrieval method that improves retrieval performance by using sublimated deep features. The method incorporates orientation-selective features and color perceptual features, effectively mimicking these mechanisms to provide a more discriminating representation.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Fengguang Peng, Zihan Ding, Ziming Chen, Gang Wang, Tianrui Hui, Si Liu, Hang Shi
Summary: RGB-Thermal (RGB-T) semantic segmentation is an emerging task that aims to improve the robustness of segmentation methods under extreme imaging conditions by using thermal infrared modality. The challenges of foreground-background distinguishment and complementary information mining are addressed by proposing a cross modulation process with two collaborative components. Experimental results show that the proposed method achieves state-of-the-art performances on current RGB-T segmentation benchmarks.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Baihong Han, Xiaoyan Jiang, Zhijun Fang, Hamido Fujita, Yongbin Gao
Summary: This paper proposes a novel automatic prompt generation method called F-SCP, which focuses on generating accurate prompts for low-accuracy classes and similar classes. Experimental results show that our approach outperforms state-of-the-art methods on six multi-domain datasets.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Huikai Liu, Ao Zhang, Wenqian Zhu, Bin Fu, Bingjian Ding, Shengwu Xiong
Summary: Adverse weather conditions present challenges for computer vision tasks, and image de-weathering is an important component of image restoration. This paper proposes a multi-patch skip-forward structure and a Residual Deformable Convolutional module to improve feature extraction and pixel-wise reconstruction.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Oliver M. Crook, Mihai Cucuringu, Tim Hurst, Carola-Bibiane Schonlieb, Matthew Thorpe, Konstantinos C. Zygalakis
Summary: The transportation LP distance (TLP) is a generalization of the Wasserstein WP distance that can be applied directly to color or multi-channelled images, as well as multivariate time-series. TLP interprets signals as functions, while WP interprets signals as measures. Although both distances are powerful tools in modeling data with spatial or temporal perturbations, their computational cost can be prohibitively high for moderate pattern recognition tasks. The linear Wasserstein distance offers a method for projecting signals into a Euclidean space, and in this study, we propose linear versions of the TLP distance (LTLP) that show significant improvement over the linear WP distance in signal processing tasks while being several orders of magnitude faster to compute than the TLP distance.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Haitao Tian, Shiru Qu, Pierre Payeur
Summary: This paper proposes a method of target-dependent classifier, which optimizes the joint hypothesis of domain adaptation into a target-dependent hypothesis that better fits with the target domain clusters through an unsupervised fine-tuning strategy and the concept of meta-learning. Experimental results demonstrate that this method outperforms existing techniques in synthetic-to-real adaptation and cross-city adaptation benchmarks.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Qingsen Yan, Axi Niu, Chaoqun Wang, Wei Dong, Marcin Wozniak, Yanning Zhang
Summary: Deep learning-based methods have achieved remarkable results in the field of super-resolution. However, the limitation of paired training image sets has led researchers to explore self-supervised learning. However, the assumption of inaccurate downscaling kernel functions often leads to degraded results. To address this issue, this paper introduces KGSR, a kernel-guided network that trains both upscaling and downscaling networks to generate high-quality high-resolution images even without knowing the actual downscaling process.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Yifan Chen, Xuelong Li
Summary: Gait recognition is a popular technology for identification due to its ability to capture gait features over long distances without cooperation. However, current methods face challenges as they use a single network to extract both temporal and spatial features. To solve this problem, we propose a two-branch network that focuses on spatial and temporal feature extraction separately. By combining these features, we can effectively learn the spatio-temporal information of gait sequences.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Wei Shi, Wentao Zhang, Wei-shi Zheng, Ruixuan Wang
Summary: This article proposes a simple yet effective visualization framework called PAMI, which does not require detailed model structure and parameters to obtain visualization results. It can be applied to various prediction tasks with different model backbones and input formats.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Xiaobo Hu, Jianbo Su, Jun Zhang
Summary: This paper reviews the latest technologies in pattern recognition, highlighting their instabilities and failures in practical applications. From a control perspective, the significance of disturbance rejection in pattern recognition is discussed, and the existing problems are summarized. Finally, potential solutions related to the application of compensation on features are discussed to emphasize future research directions.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Andres Felipe Posada-Moreno, Nikita Surya, Sebastian Trimpe
Summary: Convolutional neural networks are widely used in critical systems, and explainable artificial intelligence has proposed methods for generating high-level explanations. However, these methods lack the ability to determine the location of concepts. To address this, we propose a novel method for automatic concept extraction and localization based on pixel-wise aggregations, and validate it using synthetic datasets.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Peng Bao, Jianian Li, Rong Yan, Zhongyi Liu
Summary: In this paper, a novel Dynamic Graph Contrastive Learning framework, DyGCL, is proposed to capture the temporal consistency in dynamic graphs and achieve good performance in node representation learning.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Kristian Schultz, Saptarshi Bej, Waldemar Hahn, Markus Wolfien, Prashant Srivastava, Olaf Wolkenhauer
Summary: Research indicates that deep generative models perform poorly compared to linear interpolation-based methods for synthetic data generation on small, imbalanced tabular datasets. To address this, a new approach called ConvGeN, combining convex space learning with deep generative models, has been proposed. ConvGeN improves imbalanced classification on small datasets while remaining competitive with existing linear interpolation methods.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Khondaker Tasrif Noor, Antonio Robles-Kelly
Summary: In this paper, the authors propose H-CapsNet, a capsule network designed for hierarchical image classification. The network effectively captures hierarchical relationships using dedicated capsules for each class hierarchy. A modified hinge loss is utilized to enforce consistency among the involved hierarchies. Additionally, a strategy for dynamically adjusting training parameters is presented to achieve better balance between the class hierarchies. Experimental results demonstrate that H-CapsNet outperforms competing hierarchical classification networks.
PATTERN RECOGNITION
(2024)
Article
Computer Science, Artificial Intelligence
Lei Liu, Guorun Li, Yuefeng Du, Xiaoyu Li, Xiuheng Wu, Zhi Qiao, Tianyi Wang
Summary: This study proposes a new agricultural image segmentation model called CS-Net, which uses Simple-Attention Block and Simpleformer to improve accuracy and inference speed, and addresses the issue of performance collapse of Transformers in agricultural image processing.
PATTERN RECOGNITION
(2024)
