Article
Computer Science, Artificial Intelligence
Wen Yu, Baiying Lei, Michael K. Ng, Albert C. Cheung, Yanyan Shen, Shuqiang Wang
Summary: The proposed model in this study combines deep learning and high-order pooling techniques for early diagnosis of AD, and demonstrates superior performance in MRI image classification.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Rongqiang Tang, Housheng Su, Yi Zou, Xinsong Yang
Summary: This article introduces a new control scheme for finite-time synchronization of Markovian neural networks with time-varying delays and intermittent quantized controller. By using a novel finite-time stability inequality and linear programming method, sufficient conditions are obtained to ensure synchronization with an isolated node within a specific settling time, considering factors such as initial values, control intervals, rest intervals, and time delays. Control gains are designed through LP, and an optimal algorithm is provided to improve the accuracy of estimating settling time. A numerical example is presented to demonstrate the advantages and correctness of the theoretical analysis.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Damien Dablain, Bartosz Krawczyk, Nitesh Chawla
Summary: In this study, we propose a novel oversampling algorithm called DeepSMOTE for deep learning models, which generates high-quality artificial images to increase the number of samples for minority classes and balance the training set.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Maolong Lv, Wenwu Yu, Jinde Cao, Simone Baldi
Summary: This work proposes a reduced-complexity adaptive methodology for consensus tracking in a team of uncertain high-order nonlinear systems with switched dynamics. By defining separable functions and formulating a separation-based lemma, complexity in control design is effectively reduced.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Li Zhong, Zhen Fang, Feng Liu, Bo Yuan, Guangquan Zhang, Jie Lu
Summary: This article discusses the problem of handling unknown classes in unsupervised open set domain adaptation (UOSDA). A new upper bound risk function is proposed, and a solution is presented for the issue of open set difference in deep neural networks (DNNs). Source-domain risk and epsilon-open set difference are minimized through gradient descent, while distributional discrepancy is minimized using a novel adversarial training strategy. Experimental results show that the proposed method achieves state-of-the-art performance on multiple datasets.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Qinbin Li, Yiqun Diao, Quan Chen, Bing Sheng He
Summary: Due to increasing privacy concerns and data regulations, training data has become more fragmented, forming distributed databases of multiple data silos. Federated learning (FL) has emerged as a solution to train machine learning models collaboratively without exchanging raw data. However, the heterogeneity of data distribution among parties is a common challenge in distributed databases. Previous studies lack a comprehensive understanding of the advantages and disadvantages of FL algorithms under non-IID data settings. This paper proposes comprehensive data partitioning strategies and conducts extensive experiments to evaluate state-of-the-art FL algorithms, providing insights for future studies in addressing challenges in data silos.
2022 IEEE 38TH INTERNATIONAL CONFERENCE ON DATA ENGINEERING (ICDE 2022)
(2022)
Review
Computer Science, Theory & Methods
Davinder Kaur, Suleyman Uslu, Kaley J. Rittichier, Arjan Durresi
Summary: Artificial intelligence and algorithmic decision making have a significant impact on our daily lives, but they can sometimes cause harm to users and society. Ensuring the safety, reliability, and trustworthiness of these systems is essential. This survey examines various requirements for building trustworthy AI systems, discusses strategies for validating and verifying these systems, and provides an overview of recent advancements in the field.
ACM COMPUTING SURVEYS
(2023)
Review
Computer Science, Software Engineering
Xin Ning, Xinran Wang, Shaohui Xu, Weiwei Cai, Liping Zhang, Lina Yu, Wenfa Li
Summary: This article summarizes the recent research on Co-training algorithm, which is a main method of semi-supervised learning in machine learning. It introduces the main steps of relevant Co-training algorithms and discusses the existing problems. Suggestions for improvement and future development directions are also provided.
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
(2023)
Article
Computer Science, Artificial Intelligence
Lu Li, Wei Li, Ying Qu, Chunhui Zhao, Ran Tao, Qian Du
Summary: This article proposes a prior-based tensor approximation method for hyperspectral anomaly detection, which combines low-rank prior and piecewise-smooth prior into the background tensor, and incorporates spatial group sparse prior into the anomaly tensor, achieving effective anomaly detection.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Kejie Xu, Hong Huang, Peifang Deng, Yuan Li
Summary: A novel deep feature aggregation framework driven by graph convolutional network is developed for scene classification of high spatial resolution images. Experimental results demonstrate that the proposed method achieves more competitive performance in terms of OAs compared to some state-of-the-art methods in scene classification.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Information Systems
Juntao Chen, Quan Zou, Jing Li
Summary: N6-methyladenosine (m(6)A) is a prevalent methylation modification that is related to common diseases such as cancer, tumors, and obesity. Accurate prediction of m(6)A methylation sites in RNA sequences has become a critical issue in bioinformatics. Researchers developed an m(6)A site predictor called DeepM6ASeq-EL, which integrates LSTM and CNN classifiers with the strategy of hard voting. However, its accuracy in m(6)A site prediction is lower compared to the state-of-the-art method WHISTLE.
FRONTIERS OF COMPUTER SCIENCE
(2022)
Article
Computer Science, Information Systems
Zhihan Lv, Dongliang Chen, Haibin Lv
Summary: This article explores the method of using digital twins and BIM to process big data, in order to accelerate the construction of smart cities and improve the accuracy of data processing. By using multiple GPUs and the Bayesian network structural learning algorithm, complex data in smart cities can be effectively processed.
ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Jing Jin, Zhiqiang Wang, Ren Xu, Chang Liu, Xingyu Wang, Andrzej Cichocki
Summary: The steady-state visual evoked potential (SSVEP)-based brain-computer interface (BCI) has been widely studied due to its advantages in training time, recognition performance, and information transmission rate. This article introduces a novel time filter and similarity measurement methods based on task-related component analysis (TRCA) to improve the detection ability of SSVEPs. Experimental results demonstrate that the proposed methods outperform existing methods and show promising potential for SSVEP detection.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Jize Xue, Yongqiang Zhao, Shaoguang Huang, Wenzhi Liao, Jonathan Cheung-Wai Chan, Seong G. Kong
Summary: This paper introduces a new multilayer sparsity-based tensor decomposition method for low-rank tensor completion. By encoding the structured sparsity of a tensor through multiple-layer representation and introducing a new sparsity insight concept, it achieves a refined description of factor/subspace sparsity.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Yiyang Zhang, Feng Liu, Zhen Fang, Bo Yuan, Guangquan Zhang, Jie Lu
Summary: This article introduces a new method in unsupervised domain adaptation using complementary label data, provides a theoretical bound, and considers two different scenarios. By proposing the complementary label adversarial network CLARINET, the CC-UDA and PC-UDA problems are solved, showing superior performance in experiments.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Qi Wang, Wei Huang, Zhitong Xiong, Xuelong Li
Summary: This article proposes a multiscale representation method for scene classification in remote sensing images, which is achieved through a global-local two-stream architecture, and achieved good results.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Jin-Fan Hu, Ting-Zhu Huang, Liang-Jian Deng, Tai-Xiang Jiang, Gemine Vivone, Jocelyn Chanussot
Summary: The paper introduces a deep convolutional neural network architecture to fuse low-resolution HSI and high-resolution multispectral image for generating high-resolution HSI. By preserving spatial and spectral information using LR-HSI and HR-MSI, and utilizing attention and pixelShuffle modules for high-quality spatial details extraction, the proposed network achieves the best performance compared to recent HSI super-resolution approaches.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Theory & Methods
Jie Feng, Lei Liu, Qingqi Pei, Keqin Li
Summary: This article introduces a cost optimization problem in federated learning to ensure the convergence rate in terms of cost in wireless edge networks. By decomposing the problem into sub-problems and utilizing different algorithms to solve them, the proposed scheme achieves a tradeoff between cost and fairness.
IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS
(2022)
Article
Computer Science, Information Systems
Arman Haghanifar, Mahdiyar Molahasani Majdabadi, Younhee Choi, S. Deivalakshmi, Seokbum Ko
Summary: This study focuses on efficiently detecting imaging features of novel coronavirus pneumonia using deep convolutional neural networks. The proposed COVID-CXNet model is capable of precise localization based on relevant and meaningful features, which is a step towards a fully automated and robust COVID-19 detection system.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Yonghua Zhu, Junbo Ma, Changan Yuan, Xiaofeng Zhu
Summary: This paper proposes a GCN architecture combining interpretable feature learning and dynamic graph learning for personalized early Alzheimer's disease diagnosis, which outputs competitive diagnosis performance and provides interpretability.
INFORMATION FUSION
(2022)
