Article
Computer Science, Artificial Intelligence
HyunJin Kim
Summary: AresB-Net introduces a novel network model for accurately binarized residual convolutional neural networks, reducing errors from binarization through feature reuse and decreasing computational costs and storage requirements. Despite low hardware costs, the model achieves remarkable classification accuracies on CIFAR and ImageNet datasets.
PEERJ COMPUTER SCIENCE
(2021)
Article
Computer Science, Artificial Intelligence
HyunJin Kim, Mohammed Alnemari, Nader Bagherzadeh
Summary: This paper proposes a storage-efficient ensemble classification method to improve the classification accuracy of binary neural networks by sharing filters from a trained convolutional neural network model. Experimental results show that the method demonstrates high scalability and effectiveness on CIFAR datasets.
PEERJ COMPUTER SCIENCE
(2022)
Article
Computer Science, Artificial Intelligence
Giorgos Bouritsas, Fabrizio Frasca, Stefanos Zafeiriou, Michael M. Bronstein
Summary: Although Graph Neural Networks (GNNs) have achieved remarkable results, recent studies have shown important shortcomings in their ability to capture the structure of the underlying graph. We propose Graph Substructure Networks (GSN), a topologically-aware message passing scheme based on substructure encoding, to address these limitations and obtain state-of-the-art results in various real-world settings.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Computer Science, Information Systems
Hyunwoo Kim, Jeonghoon Kim, Jungwook Choi, Jungkeol Lee, Yong Ho Song
Summary: In this paper, a binarized encoder-decoder network (BEDN) and a binarized deconvolution engine (BiDE) are proposed to accelerate low-power, real-time semantic segmentation, reducing hardware resource usage and increasing network throughput. BiDE achieves a performance of 1.682 Tera operations per second (TOPS) and 824 Giga operations per second per watt (GOPS/W) in processing CamVid11 images at 25.89 frames per second (FPS).
Article
Computer Science, Information Systems
Xiaojun Kang, Xinchuan Li, Hong Yao, Dan Li, Bo Jiang, Xiaoyue Peng, Tiejun Wu, Shihua Qi, Lijun Dong
Summary: The technique of graph/network embedding is used to analyze and process complex graph data efficiently, and the Dynamic Hypergraph Neural Networks based on Key Hyperedges (DHKH) model is proposed to address the information transmission issue in static hypergraph structure.
INFORMATION SCIENCES
(2022)
Article
Optics
Long Huang, Jianping Yao
Summary: We propose and experimentally demonstrate an optical processor for a binarized neural network, which is capable of implementing positive and negative weights as well as multiply-accumulate operations. The accumulation operation is achieved through dispersion-induced time delays and detection at a photodetector. A proof-of-concept experiment shows the high speed and large bandwidth parallel processing capability of the processor in binarized convolutional neural network tasks.
Article
Computer Science, Artificial Intelligence
Jungwoo Shin, HyunJin Kim
Summary: PresB-Net is a novel performance-enhancing binarized neural network model that combines several state-of-the-art structures and introduces a new normalization approach. Achieving 73.84% Top-1 inference accuracy on the CIFAR-100 dataset, PresB-Net-18 outperforms other existing counterparts.
PEERJ COMPUTER SCIENCE
(2022)
Article
Computer Science, Artificial Intelligence
Babatounde Moctard Oloulade, Jianliang Gao, Jiamin Chen, Raeed Al-Sabri, Tengfei Lyu
Summary: This paper proposes a performance predictor-based graph neural architecture search (PGNAS) framework, which consists of three conceptually simpler and basic phases and can explore a search space with a cheaper computation cost. Experimental results show that PGNAS outperforms both handcrafted and Graph-NAS models on four benchmark datasets.
Article
Engineering, Electrical & Electronic
Hyeongsu Kim, Sung Yun Woo, Soochang Lee, Young-Tak Seo, Byung-Gook Park, Jong-Ho Lee
Summary: This letter proposes a variation-tolerant capacitive binarized neural network using 2-terminal MANOS memory devices. The capacitance-voltage characteristic of the synaptic device is used to represent the weight of the binarized neural network. The proposed synaptic device shows less capacitance variation compared to conventional synaptic devices, resulting in a more robust network.
IEEE ELECTRON DEVICE LETTERS
(2022)
Article
Computer Science, Artificial Intelligence
Zhenhua Huang, Wenhao Zhou, Kunhao Li, Zhaohong Jia
Summary: Graph neural networks have achieved great success in graph processing, but have limitations in feature aggregations and update mechanisms. To address these limitations, researchers propose a scalable graph convolution network (SGCN) with an expressive feature aggregation mechanism to enhance structural information learning.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Hardware & Architecture
Yiyang Jiang, Fan Yang, Bei Yu, Dian Zhou, Xuan Zeng
Summary: Layout hotspot detection is crucial in the physical verification flow, and deep neural network models have shown great success in this area. This article introduces a new deep learning architecture based on binarized neural networks for speeding up hotspot detection. Experimental results demonstrate that the proposed architecture outperforms previous hotspot detectors in accuracy and is 8 times faster than the best deep learning-based solution.
IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Desheng Wu, Quanbin Wang, David L. Olson
Summary: The increasing number and trade volume of Chinese firms have posed challenges for risk control and government supervision. This study utilizes hidden information from the supply chain network to classify participating companies, showing the effectiveness and economic significance of the graph neural network (GNN) algorithm.
APPLIED SOFT COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Jian Gao, Jianshe Wu
Summary: The high complexity of graph neural networks (GNNs) on large-scale networks hinders their industrial application. Graph condensation (GCond) and Multiple Sparse Graphs Condensation (MSGC) are proposed to address this problem. MSGC, which condenses the original large-scale graph into multiple small-scale sparse graphs, allows GNNs to obtain numerous sets of embeddings, significantly enriching the diversity of embeddings. Experimental results show that MSGC has significant advantages over GCond and other baselines at the same condensed graph scale.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Information Systems
Xianhang Zhang, Hanchen Wang, Jianke Yu, Chen Chen, Xiaoyang Wang, Wenjie Zhang
Summary: This paper proposes a bipartite graph-based capsule network, called Bipartite Capsule Graph Neural Network (BCGNN), for bipartite graph classification task. BCGNN utilizes the capsule network and obtains information between the same type vertices in the bipartite graphs by constructing the one-mode projection. Extensive experiments on real-world datasets are conducted to demonstrate the effectiveness of the proposed method.
WORLD WIDE WEB-INTERNET AND WEB INFORMATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Kaixuan Chen, Jie Song, Shunyu Liu, Na Yu, Zunlei Feng, Gengshi Han, Mingli Song
Summary: This article introduces a new pooling module called Distribution Knowledge Embedding (DKEPool) to address the issue of information missing caused by existing averaging or summing operations. By incorporating both graph structure and node sampling distribution, the proposed DKEPool can effectively summarize the entire graph information and improve the performance of graph classification tasks.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Hardware & Architecture
Kai Wang, Xuemin Lin, Lu Qin, Wenjie Zhang, Ying Zhang
Summary: This paper investigates the problem of efficiently counting butterflies in bipartite graphs and proposes a vertex-priority-based algorithm BFC-VP along with cache-aware strategies for external and parallel contexts. It also tackles the butterfly counting problem on batch-dynamic graphs with fast vertex-priority-based algorithms and optimizations to reduce computation. Extensive empirical studies show that the proposed techniques outperform baseline solutions on real datasets.
Article
Computer Science, Artificial Intelligence
Kai Wang, Gengda Zhao, Wenjie Zhang, Xuemin Lin, Ying Zhang, Yizhang He, Chunxiao Li
Summary: In this article, we propose the (alpha,beta,eta)-core model for uncertain bipartite graphs and present algorithms and index construction methods. The efficiency and effectiveness of our proposed techniques are validated through extensive experiments.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Allergy
Changyi Xu, Lijuan Du, Yubiao Guo, Yuxia Liang
Summary: This study aimed to identify differentially expressed circRNAs in induced sputum cells of asthma patients, in order to provide potential biomarkers and insights for asthma research. Through high-throughput sequencing, the differentially expressed circRNAs in asthma patients were screened, and their expression was verified by qRT-PCR. The correlation between circRNA and asthma was analyzed, and a possible ceRNA network was predicted and analyzed.
INTERNATIONAL ARCHIVES OF ALLERGY AND IMMUNOLOGY
(2023)
Article
Computer Science, Hardware & Architecture
Dian Ouyang, Dong Wen, Lu Qin, Lijun Chang, Xuemin Lin, Ying Zhang
Summary: This paper addresses the issue of computing the shortest distance between two vertices in road networks. The authors propose a novel hierarchical 2-hop index (H2H-Index) and a query processing algorithm based on this index to overcome the limitations of existing solutions. Experimental results show that their approach achieves a significant speedup in query processing compared to state-of-the-art methods.
Article
Computer Science, Interdisciplinary Applications
Chengpei Xu, Wenjing Jia, Ruomei Wang, Xiangjian He, Baoquan Zhao, Yuanfang Zhang
Summary: With the lack of elaborating annotations and interesting content in educational videos, this article proposes a slide-based video navigation tool that extracts the hierarchical structure and semantic relationship of visual entities in videos by integrating multichannel information. Through a novel deep learning framework, features of visual entities are extracted from presentation slides, and a clustering approach is used to determine the hierarchical relationships between these entities. By evaluating their semantic relationships, visual entities are associated with their corresponding audio speech text, generating a multilevel table of contents and notes for improved learning navigation.
IEEE TRANSACTIONS ON LEARNING TECHNOLOGIES
(2023)
Article
Computer Science, Information Systems
Muhammad Babar, Mian Ahmad Jan, Xiangjian He, Muhammad Usman Tariq, Spyridon Mastorakis, Ryan Alturki
Summary: With the rise of IoT, the awareness of edge computing is gaining importance. However, edge computing faces challenges in tackling the diverse applications of IoT due to the massive heterogeneous data they produce. To address these challenges, we propose an optimized IoT-enabled big data analytics architecture for edge-cloud computing using machine learning.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Zi Chen, Long Yuan, Xuemin Lin, Lu Qin, Wenjie Zhang
Summary: The existing clique model is inapplicable for signed networks, so a balanced clique model is proposed. The maximal balanced clique enumeration problem and the maximum balanced clique search problem are studied, and solutions are proposed. Extensive experiments demonstrate the efficiency, effectiveness and scalability of the algorithms.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Biology
Zihao Sun, Xiujing Chen, Xiaoning Huang, Yanfen Wu, Lijuan Shao, Suna Zhou, Zhu Zheng, Yiguang Lin, Size Chen
Summary: This study investigated the impact of cuproptosis-associated immune-related genes on lung adenocarcinoma (LUAD). The researchers identified three key genes and established a risk model and nomogram to predict the immunotherapy response and prognosis of LUAD.
Article
Computer Science, Artificial Intelligence
Yuhang Su, Yu Gu, Zhigang Wang, Ying Zhang, Jianbin Qin, Ge Yu
Summary: This paper studies the subhypergraph matching problem and proposes an efficient solution with two novel techniques. The proposed method outperforms existing methods on both real and synthetic data sets.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Lijun Chang, Xing Feng, Kai Yao, Lu Qin, Wenjie Zhang
Summary: Computing the graph edit distance (GED) is important in graph similarity search. The existing index structures are ineffective in reducing processing time, so verifying GED directly is still the best option. However, the AStar-LSa algorithm may consume a lot of memory, but we propose a new estimation and efficient algorithms to improve efficiency.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Xubo Wang, Dong Wen, Lu Qin, Lijun Chang, Ying Zhang, Wenjie Zhang
Summary: Designing distributed graph systems has attracted a lot of attention due to the expressive graph model and increasing graph volume. This paper introduces ScaleG, a novel disk-based distributed graph system, that provides user-friendly programming interfaces and reduces disk and network communication for high computational efficiency.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Kai Wang, Wenjie Zhang, Ying Zhang, Lu Qin, Yuting Zhang
Summary: This paper investigates the significant (alpha,beta)-community search problem on weighted bipartite graphs. The goal is to find the significant (alpha,beta)-community R of a query vertex q that characterizes the engagement level of vertices using (alpha,beta)-core, and maximizes the minimum edge weight (significance) within R. To support fast retrieval, a novel index structure and efficient index maintenance techniques are proposed. Peeling and expansion algorithms are developed to obtain R. Experimental results validate the effectiveness and efficiency of the proposed techniques.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)