A5100429133- Advanced Graph Neural Networks
- Privacy-Preserving Technologies in Data
- Complex Network Analysis Techniques
- Recommender Systems and Techniques
- Blockchain Technology Applications and Security
- Tensor decomposition and applications
- Sparse and Compressive Sensing Techniques
- Stochastic Gradient Optimization Techniques
- Face and Expression Recognition
- Cerebrovascular and Carotid Artery Diseases
- Caching and Content Delivery
- Intracranial Aneurysms: Treatment and Complications
- Advanced Computational Techniques and Applications
- Anomaly Detection Techniques and Applications
- Advanced Clustering Algorithms Research
- Network Security and Intrusion Detection
- Topic Modeling
- Service-Oriented Architecture and Web Services
- Internet Traffic Analysis and Secure E-voting
- Acute Ischemic Stroke Management
- Domain Adaptation and Few-Shot Learning
- Cryptography and Data Security
- Opinion Dynamics and Social Influence
- Advanced Neuroimaging Techniques and Applications
- Video Surveillance and Tracking Methods
Sun Yat-sen University
2016-2025
China West Normal University
2021-2025
Beijing Forestry University
2025
Technical University of Munich
2023-2025
Helix (United States)
2025
Xinjiang University
2018-2025
University of Pittsburgh
2020-2024
Zhejiang University of Technology
2024
BGI Group (China)
2023-2024
BGI Research
2023-2024
Federated learning has been widely studied and applied to various scenarios, such as financial credit, medical identification, so on. Under these settings, federated protects users from exposing their private data, while cooperatively training a shared machine algorithm model (i.e., the global model) for variety of realworld applications. The only data exchanged is gradient or updated local update). However, security increasingly being questioned, due malicious clients central servers'...
Community structure is ubiquitous in real-world complex networks. The task of community detection over these networks paramount importance a variety applications. Recently, nonnegative matrix factorization (NMF) has been widely adopted for due to its great interpretability and natural fitness capturing the membership nodes. However, existing NMF-based approaches are shallow methods. They learn assignment by mapping original network space directly. Considering complicated diversified topology...
In this paper, we propose a secure and efficient blockchain-based data trading approach for the Internet of Vehicles (IoV). First, apply consortium blockchain technologies to ensure truthful trading, general framework IoV. Second, improve efficiency encourage more participants trade data, an iterative double auction mechanism with purpose achieving social welfare maximization, in which pricing rules buyers sellers are designed induce submit bids decide amount traded its price among sellers....
With the continuous development of web technology, Web 3.0 has attracted a considerable amount attention due to its unique decentralized characteristics. The digital economy is an important driver high-quality economic and currently in rapid stage. In scenario, centralized nature Internet other characteristics usually bring about security issues such as infringement privacy leakage. Therefore, it necessary investigate how use technologies solve pain points encountered by fully exploring...
Disinfection is an indispensable process in wastewater treatment plants. New bacterial inactivation technologies are of increasing interest and persistent demand. A category simple efficient bactericidal systems have been established this study, that is, the combination divalent transition metal (Mn(II), Co(II), Fe(II), or Cu(II)) sulfite. In these systems, catalyzed auto-oxidation sulfite was manifested to generate reactive intermediary SO4•– played major role Escherichia coli at pH 5–8.5....
Offloading computation-intensive blockchain mining tasks to the edge servers (ESs) is a promising solution for blockchain-empowered Industrial Internet of Things (IIoT) because computing capabilities in IIoT are usually limited, whereas computationally intensive. However, computation offloading solutions data processing and have been studied separately. Moreover, most existing assume that all devices can directly connect ESs or cloud centers. To address these issues, this paper, we propose...
The cryptocurrency market is a very huge without effective supervision. It of great importance for investors and regulators to recognize whether there are manipulation its patterns. This paper proposes an approach mine the transaction networks exchanges answering this question. By taking leaked history Mt. Gox Bitcoin exchange as sample, we first divide accounts into three categories according characteristic then construct graphs. Many observations findings obtained via analyzing constructed...
Recently, blockchain technology has become a topic in the spotlight but also hotbed of various cybercrimes. Among them, phishing scams on have been found to make notable amount money, thus emerging as serious threat trading security ecosystem. In order create favorable environment for investment, an effective method detecting is urgently needed To this end, article proposes approach detect Ethereum by mining its transaction records. Specifically, we first crawl labeled addresses from two...
Federated learning allows multiple participants to collaboratively train an efficient model without exposing data privacy. However, this distributed machine training method is prone attacks from Byzantine clients, which interfere with the of global by modifying or uploading false gradient. In paper, we propose a novel serverless federated framework Committee Mechanism based Learning (CMFL), can ensure robustness algorithm convergence guarantee. CMFL, committee system set up screen uploaded...
Mammographic density (MD) has not been systematically investigated among Chinese women. Breast cancer screening programmes provided detailed information on MD in a large number of asymptomatic women.In the Multi-modality Independent Screening Trial (MIST), we estimated association between and its influential factors using logistic regression, adjusting for age, body mass index (BMI) study area. Differences other ethnic groups with respect to were also explored adjustment age BMI.A total 28...
In this paper, we study an eigenvalue problem for even order tensors. Using the matrix unfolding of tensors, can establish relationship between a tensor and multilevel problem. By considering higher singular value decomposition tensor, show that values are square root eigenvalues product its conjugate transpose. This result is similar to in case. Also Toeplitz/circulant give lower upper bounds Toeplitz An application image restoration also discussed.
Abstract Patients with chronic obstructive pulmonary disease (COPD) repeat acute exacerbations (AE). Global Initiative for Chronic Obstructive Lung Disease (GOLD) is only available patients in stable phase. Currently, there a lack of assessment and prediction methods exacerbation (AECOPD) during hospitalization. To enhance the monitoring treatment AECOPD patients, we develop novel C5.0 decision tree classifier to predict prognosis hospitalized objective clinical indicators. The medical...
Intrauterine adhesions (IUA), which is characterized by endometrial fibrosis, continue to be the most common cause of uterine infertility globally. Our work revealed that 3 fibrotic progression markers (Vimentin, COL5A2, and COL1A1) were significantly increased in endometrium IUA patients. Mesenchymal stem cell-derived exosomes (EXOs) have been recently as a cell-free therapy for fibrosis diseases. Nevertheless, application EXOs restricted short residency duration target tissue. To overcome...
Avatars, as promising digital assistants in Vehicular Metaverses, can enable drivers and passengers to immerse 3-D virtual spaces, serving a practical emerging example of Artificial Intelligence Things (AIoT) intelligent vehicular environments. The immersive experience is achieved through seamless human–avatar interaction, e.g., augmented reality navigation, which requires intensive resources that are inefficient impractical process on vehicles locally. Fortunately, offloading avatar tasks...
Recently, nonnegative matrix factorization (NMF) has been widely adopted for community detection, because of its better interpretability. However, the existing NMF-based methods have following three problems: 1) they directly transform original network into membership space, so it is difficult them to capture hierarchical information; 2) often only pay attention topology and ignore node attributes; 3) hard learn global structure information necessary detection. Therefore, we propose a new...
The increasing necessity to establish a permanent presence in overseas markets challenges the tradition of tide‐like market entry mode international contractors and confronts them with an important dichotomous selection between mobile mode. By borrowing theories previous findings from general business area, hypotheses regarding influences specific host country related factors upon two generic modes, including cultural difference, trade link, potential, investment risk, institutional barrier...
Arsenic species transformation in paddy soils has important implications for arsenic accumulation rice grains and its safety to the consumers. Methylated thioarsenates including highly toxic dimethylated monothioarsenate (DMMTA) have been detected soils, but their production dynamics remain poorly understood. In present study, we first optimized a HPLC-ICP-MS method quantify methylated thioarsenate species. Using this together with 10 mM diethylenetriamine pentaacetate (DTPA) preserve As...
In this paper, we address the total variation (TV)-based nonlinear image restoration problems. problems, an original is corrupted by a spatially-invariant blur, build-in nonlinearity in imaging system, and additive Gaussian white noise. We study objective function consisting of least squares data-fitting term TV regularization restored image. By making use structure function, efficient alternating direction method multipliers can be developed for solving proposed model. The convergence...