A5100377070- Advanced Graph Neural Networks
- Topic Modeling
- IoT and Edge/Fog Computing
- Recommender Systems and Techniques
- Adversarial Robustness in Machine Learning
- Software-Defined Networks and 5G
- Advanced Steganography and Watermarking Techniques
- Advanced Computational Techniques and Applications
- Explainable Artificial Intelligence (XAI)
- Machine Learning and Data Classification
- Advanced Vision and Imaging
- Advanced Neural Network Applications
- Opportunistic and Delay-Tolerant Networks
- Blockchain Technology Applications and Security
- Data Visualization and Analytics
- Chaos-based Image/Signal Encryption
- Distributed systems and fault tolerance
- Computer Graphics and Visualization Techniques
- Advanced Memory and Neural Computing
- Cell Image Analysis Techniques
- Parallel Computing and Optimization Techniques
- Network Security and Intrusion Detection
- Natural Language Processing Techniques
- IoT Networks and Protocols
- Biometric Identification and Security
Henan University of Science and Technology
2025
Nanjing Traditional Chinese Medicine Hospital
2024
University of Illinois Chicago
2016-2023
North China Electric Power University
2023
Aisino (China)
2018-2022
Amazon (United States)
2022
Franklin & Marshall College
2022
Google (United States)
2021
Zhejiang University of Science and Technology
2017-2019
Software (Spain)
2019
Massive open online courses (MOOCs) are becoming a modish way for education, which provides large-scale and open-access learning opportunity students to grasp the knowledge. To attract students' interest, recommendation system is applied by MOOCs providers recommend students. However, as course usually consists of number video lectures, with each one covering some specific knowledge concepts, directly recommending overlook interest concepts. fill this gap, in paper, we study problem concept...
Equipped with easy-to-access micro computation access points, the fog computing architecture provides low-latency and ubiquitously available offloading services to many simple cheap Internet of Things devices limited energy resources. One obstacle, however, is how seamlessly hand over mobile IoT among different points when in action so that service not interrupted -- especially for time-sensitive applications. In this article, we propose Follow Me Fog (FMF), a framework supporting new...
The cyber-physical system (CPS) is a promising technique that enables safety-critical industry ecosystem. In general, wireless sensor networks (WSNs) and the Internet of Things are sensing communication infrastructures for CPS. Currently, software-defined networking (SDN) have been used as new architecture typical WSNs However, there two unresolved problems First, feasible systemic must CPS, which should provide global virtualization management closed-loop control between cyber side physical...
The existing eye trackers typically require an explicit personal calibration procedure to estimate subject-dependent parameters. Despite efforts in simplifying the process, such a process remains unnatural and bothersome, particular for users of mobile devices. To alleviate this problem, we introduce technique that can eliminate calibration. Based on combining new with fixation prediction, proposed method performs implicit without active participation or even knowledge user. Specifically,...
Mining from neuroimaging data is becoming increasingly popular in the field of healthcare and bioinformatics, due to its potential discover clinically meaningful structure patterns that could facilitate understanding diagnosis neurological neuropsychiatric disorders. Most recent research concentrates on applying subgraph mining techniques connected brain network. However, underlying network complicated. As a shallow linear model, cannot capture highly non-linear structures, resulting...
The incidence of acute myeloid leukemia (AML) is increasing annually, and timely diagnostic treatments can substantially improve patient survival rates. AML typing traditionally relies on manual microscopy for classifying counting cells, which time-consuming, laborious, subjective. Therefore, developing a reliable automated model cell classification imperative. This study evaluated the performance five widely-used models largest publicly available bone marrow dataset (BM). However, accuracy...
Abstract A blockchain is a decentralized distributed public database. It does not have central authority to maintain this database by running cryptographic protocol with nodes. Bitcoin currently the hottest item in blockchain, and node can verify transaction content package it into block. The guarantees consistency of books through underlying consensus agreement. These algorithms are different because algorithm security assumptions from actual requirements. This paper sorts compares various...
Danilo Neves Ribeiro, Shen Wang, Xiaofei Ma, Rui Dong, Xiaokai Wei, Henghui Zhu, Xinchi Chen, Peng Xu, Zhiheng Huang, Andrew Arnold, Dan Roth. Findings of the Association for Computational Linguistics: NAACL 2022.
Graph neural network (GNN), as a powerful representation learning model on graph data, attracts much attention across various disciplines. However, recent studies show that GNN is vulnerable to adversarial attacks. How make more robust? What are the key vulnerabilities in GNN? address and defense against attacks? In this paper, we propose DefNet, an effective framework for GNNs. particular, first investigate latent every layer of GNNs corresponding strategies including dual-stage aggregation...
Nowadays, vehicle-to-grid (V2G) is a very important component for smart cities, which provide novel energy storage and scheduling approach. However, security threats currently disturb the normal operations of V2G. There are two challenges On one hand, existing schemes V2G just consider static strategy, cannot deal well with problem high dynamics advanced persistent threat in other lacks unified information modeling approach, results difficulties communications. To address above challenges,...
Nowadays, artificial intelligence technologies (e.g., deep neural networks) have been used widely in the Internet of Things (IoT) to provide smart services and sensing data processing. The evolving network even exceeds human cognitive level. However, accuracy these structures depends some extent on training data. Some well-designed generated antagonistic disturbances are sufficient deceive model when added images. Such attacks cause classifiers trained by misidentify object thus completely...
Most compilers for machine learning (ML) frameworks need to solve many correlated optimization problems generate efficient code. Current ML rely on heuristics based algorithms these one at a time. However, this approach is not only hard maintain but often leads sub-optimal solutions especially newer model architectures. Existing approaches in the literature are sample inefficient, tackle single problem, and do generalize unseen graphs making them infeasible be deployed practice. To address...
Fog computing, shifting intelligence and resources from the remote cloud to edge networks, has potential of providing low-latency for communication sensing data sources users. For objects Internet Things (IoT) cloud, it is a new trend that establish social-like relationships with each other, which efficiently brings benefits developed sociality complex environment. As fog service become more sophisticated, will convenient users share their own services, resources, via social networks....
Search-based techniques have been demonstrated effective in solving complex optimization problems that arise domain-specific compilers for machine learning (ML). Unfortunately, deploying such production is impeded by two limitations. First, prior works require factorization of a computation graph into smaller subgraphs over which search applied. This decomposition not only non-trivial but also significantly limits the scope optimization. Second, to be applied single stage compilation flow,...
Knowledge graph (KG) enhanced recommendation has demonstrated improved performance in the system (RecSys) and attracted considerable research interest. Recently literature adopted neural networks (GNNs) on collaborative knowledge built an end-to-end KG-enhanced RecSys. However, majority of these approaches have three limitations: (1) treat as a homogeneous overlook highly heterogeneous relationships among items, (2) lack design to explicitly leverage rich side information, (3) user...
Because the authentication method based on username‐password has disadvantage of easy disclosure and low reliability excess password management degrades user experience tremendously, is eager to get rid bond in order seek a new way authentication. Therefore, multifactor biometrics‐based wins favor people with advantages simplicity, convenience, high reliability. Now (especially fingerprint information) technology been extremely mature, it universally applied scenario mobile payment....
With integration between Internet-of-Things (IoT) and industrial systems, Industrial Internet of Things (IIoT) raises more efficiency security access for monitoring control. The IEEE 21451 standard, gradually accepted by manufacturers users, aims to solve the problem unified sensor communication standard networks, especially wireless networks. Although 21451-1-5 can realize simple network management IIoT, efficient during continues be an unresolved problem. In this paper, we propose secure...
This paper provides a method and system for establishing communicating cross-chain alliances between blockchains. The the alliance includes: obtaining communication requirements of each node in multiple blockchains, configuring identity certificates transaction smart contracts.The blockchain request by using established above, contract node; if is initiated, performing an endorsement policy verification on request, then full sequence signature data consistency verification, when verified,...
With the increasing of data quantity and continuous development in mining applications, introduction deep learning is imminent. However, combination still faces many challenges distributed large-scale environment, one most immediate issues to ensure privacy when running cloud. In this paper, we propose a scheme combined multi-key fully homomorphic encryption with neural networks. scheme, participates protect their which could achieve requirements preserving mining.