Article
Computer Science, Artificial Intelligence
Ronald Salloum, C. -C. Jay Kuo
Summary: This paper introduces a new data visualization and clustering technique called cPCA++, which efficiently discovers discriminative structures in high-dimensional data by highlighting dataset-specific patterns that are not detected by traditional PCA. The proposed method outperforms state-of-the-art methods in terms of efficiency and achieves similar or better performance in experiments.
PATTERN RECOGNITION
(2022)
Article
Computer Science, Information Systems
T. M. Tariq Adnan, Md Mehrab Tanjim, Muhammad Abdullah Adnan
Summary: PCA is a popular technique for dimensionality reduction, but scalability issues arise in high dimensions. This study proposes a solution based on the zero-noise-limit Probabilistic PCA model, introducing a block-division method to suppress intermediate data explosion and achieve efficient communication in a geographically distributed environment.
INFORMATION SYSTEMS
(2021)
Article
Computer Science, Information Systems
Fujiao Ju, Yanfeng Sun, Junbin Gao, Yongli Hu, Baocai Yin
Summary: Probabilistic Linear Discriminant Analysis (PLDA) as a generative model achieves good performance in supervised learning tasks by introducing Laplace prior and using Kronecker-decomposable component to enhance robustness and handle tensor data, showing significant performance advantage in experiments compared to state-of-the-art LDA-based algorithms.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Xiaoping Liang, Zhenjun Tang, Xiaolan Xie, Jingli Wu, Xianquan Zhang
Summary: The paper proposes a robust and fast image hashing based on two-dimensional (2D) principal component analysis (PCA) and saliency map, with the saliency map determined by the LC (luminance contrast) method to ensure robustness and 2D PCA used for fast and efficient dimensionality reduction to learn compact and discriminative code, providing fast speed for hashing.
MULTIMEDIA SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Elia Balugani, Francesco Lolli, Martina Pini, Anna Maria Ferrari, Paolo Neri, Rita Gamberini, Bianca Rimini
Summary: Life Cycle Assessment is quantified using Robust Ordinal Regression and Multi-Criteria Decision Analysis. Integration of Principal Component Analysis and Robust Ordinal Regression methods is proposed to reduce problem dimensionality. The effectiveness of the proposed methods is tested using generated datasets, literature datasets, and a Life Cycle Assessment case study.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Biochemical Research Methods
Tobias Wangberg, Joanna Tyrcha, Chun-Biu Li
Summary: The proposed shape-aware stochastic neighbor embedding method outperforms t-SNE, UMAP, and PHATE in visualizing imbalanced, nonlinear, continuous, and hierarchically structured data. Additionally, the method can automatically choose the optimal hyper-parameter in a data-driven way, consistently performing well across different test cases.
BMC BIOINFORMATICS
(2022)
Article
Chemistry, Multidisciplinary
Yibin Liu, Shanshan Yong, Chunjiu He, Xin'an Wang, Zhenyu Bao, Jinhan Xie, Xing Zhang
Summary: With the continuous development of technology, machine learning methods are increasingly being applied to earthquake prediction. Through big data analysis and machine learning algorithms, a new earthquake prediction method with 60% accuracy has been proposed.
APPLIED SCIENCES-BASEL
(2022)
Article
Agriculture, Multidisciplinary
Sergio Velez, Florian Rancon, Enrique Barajas, Guilhem Brunel, Jose Antonio Rubio, Bruno Tisseyre
Summary: This study utilizes Sentinel-2 satellite imagery to extract relevant information from two vineyards in Spain. By employing dimensionality reduction techniques such as Principal Component Analysis (PCA) and Partial Least Square (PLS), the NDVI time-series are decomposed into multiple functional components. The results demonstrate the added value of considering the entire time-series compared to a single image, and establish correlations with seasonal phenology and management practices in the vineyards.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Computer Science, Artificial Intelligence
Zheng Liu, Wei Jin, Ying Mu
Summary: This paper introduces a unified framework that incorporates robust graph learning and dimensionality reduction, as well as clustering task. Two robust graph methods based on Euclidean distance and self-expressiveness are proposed, which are informative, robust, and sparse. Extensive experiments demonstrate their advantages in the task of clustering.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2022)
Article
Computer Science, Theory & Methods
Nicolas Jouvin, Charles Bouveyron, Pierre Latouche
Summary: The work proposes a discriminative latent mixture model extended to Bayesian framework for clustering high-dimensional data in a low-dimensional discriminative subspace using Gaussian mixture models. Through extensive evaluation including simulations and real data, it is demonstrated that the proposed method performs well in clustering tasks.
STATISTICS AND COMPUTING
(2021)
Article
Computer Science, Software Engineering
Nana Zhang, Shi Ying, Kun Zhu, Dandan Zhu
Summary: This paper introduces a novel defect prediction model SSEPG, which combines SSDEA and ELM optimization algorithms PSO and GSA to achieve superior predictive performance. Experimental results demonstrate the superiority of SSDAE and SSEPG on multiple evaluation metrics.
Article
Computer Science, Artificial Intelligence
Yunlong Gao, Tingting Lin, Jinyan Pan, Feiping Nie, Youwei Xie
Summary: This paper proposes a new technique called Fuzzy Sparse Deviation Regularized Robust Principal Component Analysis (FSD-PCA) to improve the robustness of Principal Component Analysis (PCA) to noise samples. By introducing sparse deviation and fuzzy weighting, FSD-PCA is able to process noise and principal components separately, thus enhancing its ability to retain principal component information.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2022)
Article
Computer Science, Artificial Intelligence
Baiting Zhao, Xiao Dong, Yongcun Guo, Xiaofen Jia, Yourui Huang
Summary: A pooling method based on Principal Component Analysis (PCA) called P(CA)Pool was designed to retain more feature information and improve image classification accuracy compared to traditional pooling methods.
NEURAL PROCESSING LETTERS
(2022)
Article
Computer Science, Artificial Intelligence
Sisi Wang, Feiping Nie, Zheng Wang, Rong Wang, Xuelong Li
Summary: PCA is a powerful unsupervised dimensionality reduction algorithm that cleverly combines reconstruction error and projection variance to learn projection matrix accurately. It uses '2-norm as the evaluation criterion and has rotation invariance. It also enhances robustness and anomaly detection ability through binary weight design and employs an efficient iterative optimization algorithm to solve the problem. Extensive experimental results show that our model outperforms state-of-the-art PCA methods.
Article
Computer Science, Information Systems
Tapan Kumar Sahoo, Haider Banka, Atul Negi
Summary: This paper presents various bidirectional 2DPCAs in a feature partitioning framework, including Bi-SIMPCA, Bi-FLPCA, Bi-ESIMPCA, and Bi-EFLPCA, with optimization methods proposed for extracting 2DPCA features of images. Simulation results show that the Bi-EFLPCA method outperforms existing variations in terms of feature dimensionality, memory efficiency, recognition accuracy, and speed.
MULTIMEDIA TOOLS AND APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Tianhang Long, Yanfeng Sun, Junbin Gao, Yongli Hu, Baocai Yin
Summary: This article focuses on domain adaptation on the classic Grassmann manifolds and proposes an optimal transport-based model to achieve adaptation between datasets. It introduces regularization terms to maintain task-related consistency and presents a simplified model to reduce computational cost.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Information Systems
Jiapu Wang, Boyue Wang, Junbin Gao, Yongli Hu, Baocai Yin
Summary: This article proposes a Multi-concept Representation Learning (McRL) method for Knowledge Graph Completion (KGC), which infers missing entities or relations. Experimental results show that the proposed method outperforms existing state-of-the-art KGC methods on several standard datasets.
ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA
(2023)
Article
Engineering, Civil
Guangyu Huo, Yong Zhang, Boyue Wang, Junbin Gao, Yongli Hu, Baocai Yin
Summary: In this study, a hierarchical traffic flow forecasting network is proposed by combining the long-term temporal Transformer network (LTT) and the spatio-temporal graph convolutional networks (STGC). This method can capture both short-term and long-term temporal relations in traffic flow data and mitigate the over-smoothing problem of graph convolutional networks (GCN).
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Engineering, Biomedical
Toufique A. Soomro, Lihong Zheng, Ahmed J. Afifi, Ahmed Ali, Shafiullah Soomro, Ming Yin, Junbin Gao
Summary: Magnetic Resonance Imaging (MRI) is commonly used in brain disease detection and diagnosis, providing three-dimensional images for precise anomaly identification. However, the process is time-consuming. Machine learning and efficient computation offer a computer-aided solution for quick and accurate abnormality identification. Brain tumor segmentation from MRI images is a hot research topic, and deep learning methods are found to be more effective in this regard.
IEEE REVIEWS IN BIOMEDICAL ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Guangyu Huo, Yong Zhang, Junbin Gao, Boyue Wang, Yongli Hu, Baocai Yin
Summary: This paper proposes a cross-attention based deep clustering framework, named CaEGCN, which can extract the most discriminative information from individual data and produce more satisfactory clustering results. The framework consists of four main modules: the cross-attention fusion module, the Content Auto-encoder module (CAE), the Graph Convolutional Auto-encoder module (GAE), and the self-supervised model for clustering tasks. Experimental results demonstrate the superiority and robustness of the proposed CaEGCN.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Huidong Liang, Xingjian Du, Bilei Zhu, Zejun Ma, Ke Chen, Junbin Gao
Summary: This paper presents iGCL, an Implicit Graph Contrastive Learning method that utilizes augmentations in the latent space learned from a Variational Graph Auto-Encoder to reconstruct graph topological structure. The proposed method achieves state-of-the-art accuracy on downstream classification tasks and improves learning efficiency.
Article
Computer Science, Information Systems
Yang Li, Mingyuan Bai, Qingfeng Guan, Zi Ming, Xun Liang, Gang Liu, Junbin Gao
Summary: The reverse k nearest neighbors (RkNN) query is a time-consuming spatial query used in various domains. We propose a verification approach called CSD to reduce query time and complexity by determining whether points belong to the RkNN set. Additionally, we introduce a Voronoi-based candidate generation method to reduce the candidate set size. Experimental results demonstrate that our CSD-RkNN algorithm outperforms other algorithms, especially for large k values.
INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE
(2023)
Article
Computer Science, Artificial Intelligence
Xuebin Zheng, Bingxin Zhou, Ming Li, Yu Guang Wang, Junbin Gao
Summary: This paper proposes a framework for graph neural networks called MATHNET, which utilizes multiresolution Haar-like wavelets and interrelated convolution and pooling strategies to achieve consistent graph representations for label prediction. The proposed MATHNET outperforms other existing GNN models, particularly on large datasets.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Tengfei Liu, Yongli Hu, Boyue Wang, Yanfeng Sun, Junbin Gao, Baocai Yin
Summary: The article introduces a novel hierarchical graph convolutional network (HGCN) for structured long document classification. The network models macrostructure using a section graph network, extracts fine-grained features using a word graph network, and integrates the two networks with an interaction strategy. Experimental results show that the proposed method outperforms state-of-the-art related classification methods.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Jie Chen, Shouzhen Chen, Junbin Gao, Zengfeng Huang, Junping Zhang, Jian Pu
Summary: Due to the homophily assumption, graph neural networks (GNNs) are believed to perform well on homophilic graphs but may fail on heterophilic graphs. However, existing metrics cannot accurately explain GNNs' performance on heterophilic datasets. In this work, we propose a new metric and a framework called CAGNNs to enhance GNNs' performance on heterophily datasets. Experimental results show significant improvements compared to other GNN models.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Jiapu Wang, Boyue Wang, Junbin Gao, Xiaoyan Li, Yongli Hu, Baocai Yin
Summary: This article introduces a novel TKGC method called the quadruplet distributor network (QDN), which independently models the embeddings of entities, relations, and timestamps to fully capture the semantics and uses a quadruplet-specific decoder for integration. A novel temporal regularization method is also proposed. Experimental results demonstrate the superior performance of the proposed method.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Qi Zhang, Jinghua Li, Yanfeng Sun, Shaofan Wang, Junbin Gao, Baocai Yin
Summary: This paper proposes a novel graph neural network called AFGNN, which can capture all frequency information on large-scale graphs. AFGNN consists of two stages: the first stage extracts comprehensive frequency information using low-, middle-, and high-pass graph filters, while the second stage generates customized graph filters for each node using node-level attention-based feature combination. Experimental results demonstrate the superiority of AFGNN over other scalable GNNs and spectral GNNs.
Article
Computer Science, Artificial Intelligence
Jie Chen, Shouzhen Chen, Mingyuan Bai, Jian Pu, Junping Zhang, Junbin Gao
Summary: Graph neural networks (GNNs) are widely used in graph node classification tasks, but suffer from negative disturbances caused by edges connecting nodes with different labels. This article proposes a decoupling attention mechanism that utilizes both hard and soft attention to address this issue. Hard attention is learned on labels to refine the graph structure and reduce negative disturbances, while soft attention learns aggregation weights based on features to enhance information gains during message passing. The proposed method is validated on six benchmark graph datasets.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Jipeng Guo, Yanfeng Sun, Junbin Gao, Yongli Hu, Baocai Yin
Summary: This paper proposes a novel multi-attribute subspace clustering (MASC) model that understands data from multiple attributes. By exploiting the intrinsic multi-attribute features, MASC simultaneously learns multiple subspace representations corresponding to each specific attribute. By aggregating these multi-attribute representations, a more comprehensive subspace representation can be obtained, leading to improved clustering performance.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2022)
Article
Computer Science, Artificial Intelligence
Xiaxia He, Boyue Wang, Yongli Hu, Junbin Gao, Yanfeng Sun, Baocai Yin
Summary: In this article, we propose a novel end-to-end parallelly adaptive graph convolutional clustering (AGCC) model to improve the performance of graph convolutional network (GCN) by adaptively learning the graph structure from the data. The model consists of two pathway networks, one for updating the graph structure and the data representation layer, and the other for extracting data features. To effectively connect these two pathways, an attention-mechanism-based fusion module is also proposed.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Rui Lv, Dingheng Wang, Jiangbin Zheng, Zhao-Xu Yang
Summary: In this paper, the authors investigate tensor decomposition for neural network compression. They analyze the convergence and precision of tensor mapping theory, validate the rationality of tensor mapping and its superiority over traditional tensor approximation based on the Lottery Ticket Hypothesis. They propose an efficient method called 3D-KCPNet to compress 3D convolutional neural networks using the Kronecker canonical polyadic (KCP) tensor decomposition. Experimental results show that 3D-KCPNet achieves higher accuracy compared to the original baseline model and the corresponding tensor approximation model.
Article
Computer Science, Artificial Intelligence
Xiangkun He, Zhongxu Hu, Haohan Yang, Chen Lv
Summary: In this paper, a novel constrained multi-objective reinforcement learning algorithm is proposed for personalized end-to-end robotic control with continuous actions. The approach trains a single model using constraint design and a comprehensive index to achieve optimal policies based on user-specified preferences.
Article
Computer Science, Artificial Intelligence
Zhijian Zhuo, Bilian Chen, Shenbao Yu, Langcai Cao
Summary: In this paper, a novel method called Expansion with Contraction Method for Overlapping Community Detection (ECOCD) is proposed, which utilizes non-negative matrix factorization to obtain disjoint communities and applies expansion and contraction processes to adjust the degree of overlap. ECOCD is applicable to various networks with different properties and achieves high-quality overlapping community detection.
Article
Computer Science, Artificial Intelligence
Yizhe Zhu, Chunhui Zhang, Jialin Gao, Xin Sun, Zihan Rui, Xi Zhou
Summary: In this work, the authors propose a Contrastive Spatio-Temporal Distilling (CSTD) approach to improve the detection of high-compressed deepfake videos. The approach leverages spatial-frequency cues and temporal-contrastive alignment to fully exploit spatiotemporal inconsistency information.
Review
Computer Science, Artificial Intelligence
Laijin Meng, Xinghao Jiang, Tanfeng Sun
Summary: This paper provides a review of coverless steganographic algorithms, including the development process, known contributions, and general issues in image and video algorithms. It also discusses the security of coverless steganography from theoretical analysis to actual investigation for the first time.
Article
Computer Science, Artificial Intelligence
Yajie Bao, Tianwei Xing, Xun Chen
Summary: Visual question answering requires processing multi-modal information and effective reasoning. Neural-symbolic learning is a promising method, but current approaches lack uncertainty handling and can only provide a single answer. To address this, we propose a confidence based neural-symbolic approach that evaluates NN inferences and conducts reasoning based on confidence.
Article
Computer Science, Artificial Intelligence
Anh H. Vo, Bao T. Nguyen
Summary: Interior style classification is an interesting problem with potential applications in both commercial and academic domains. This project proposes a method named ISC-DeIT, which combines data-efficient image transformer architectures and knowledge distillation, to address the interior style classification problem. Experimental results demonstrate a significant improvement in predictive accuracy compared to other state-of-the-art methods.
Article
Computer Science, Artificial Intelligence
Shashank Kotyan, Danilo Vasconcellos Vargas
Summary: This article introduces a novel augmentation technique called Dynamic Scanning Augmentation to improve the accuracy and robustness of Vision Transformer (ViT). The technique leverages dynamic input sequences to adaptively focus on different patches, resulting in significant changes in ViT's attention mechanism. Experimental results demonstrate that Dynamic Scanning Augmentation outperforms ViT in terms of both robustness to adversarial attacks and accuracy against natural images.
Article
Computer Science, Artificial Intelligence
Hiba Alqasir, Damien Muselet, Christophe Ducottet
Summary: The article proposes a solution to improve the learning process of a classification network by providing shape priors, reducing the need for annotated data. The solution is tested on cross-domain digit classification tasks and a video surveillance application.
Article
Computer Science, Artificial Intelligence
Dexiu Ma, Mei Liu, Mingsheng Shang
Summary: This paper proposes a method using neural dynamics solvers to solve infinity-norm optimization problems. Two improved solvers are constructed and their effectiveness and superiority are demonstrated through theoretical analysis and simulation experiments.
Article
Computer Science, Artificial Intelligence
Francesco Gregoretti, Giovanni Pezzulo, Domenico Maisto
Summary: Active Inference is a computational framework that uses probabilistic inference and variational free energy minimization to describe perception, planning, and action. cpp-AIF is a header-only C++ library that provides a powerful tool for implementing Active Inference for Partially Observable Markov Decision Processes through multi-core computing. It is cross-platform and improves performance, memory management, and usability compared to existing software.
Article
Computer Science, Artificial Intelligence
Zelin Ying, Dawei Cheng, Cen Chen, Xiang Li, Peng Zhu, Yifeng Luo, Yuqi Liang
Summary: This paper proposes a novel stock market trends prediction framework called SMART, which includes a self-supervised stock technical data sequence embedding model S3E. By training with multiple self-supervised auxiliary tasks, the model encodes stock technical data sequences into embeddings and uses the learned sequence embeddings for predicting stock market trends. Extensive experiments on China A-Shares market and NASDAQ market prove the high effectiveness of our model in stock market trends prediction, and its effectiveness is further validated in real-world applications in a leading financial service provider in China.
Article
Computer Science, Artificial Intelligence
Hao Li, Hao Jiang, Dongsheng Ye, Qiang Wang, Liang Du, Yuanyuan Zeng, Liu Yuan, Yingxue Wang, C. Chen
Summary: DHGAT1, a dynamic hyperbolic graph attention network, utilizes hyperbolic metric properties to embed dynamic graphs. It employs a spatiotemporal self-attention mechanism and weighted node representations, resulting in excellent performance in link prediction tasks.
Article
Computer Science, Artificial Intelligence
Jiehui Huang, Zhenchao Tang, Xuedong He, Jun Zhou, Defeng Zhou, Calvin Yu-Chian Chen
Summary: This study proposes a progressive learning multi-scale feature blending model for image deraining tasks. The model utilizes detail dilation and texture extraction to improve the restoration of rainy images. Experimental results show that the model achieves near state-of-the-art performance in rain removal tasks and exhibits better rain removal realism.
Article
Computer Science, Artificial Intelligence
Lizhi Liu, Zilin Gao, Yinhe Wang, Yongfu Li
Summary: This paper proposes a novel discrete-time interconnected model for depicting complex dynamical networks. The model consists of nodes and edges subsystems, which consider the dynamic characteristic of both nodes and edges. By designing control strategies and coupling modes, the stabilization and synchronization of the network are achieved. Simulation results demonstrate the effectiveness of the proposed methods.