A5101454000- Recommender Systems and Techniques
- Advanced Graph Neural Networks
- Complex Network Analysis Techniques
- Network Security and Intrusion Detection
- AI in cancer detection
- Topic Modeling
- Access Control and Trust
- Service-Oriented Architecture and Web Services
- Digital Imaging for Blood Diseases
- Caching and Content Delivery
- Biomedical Text Mining and Ontologies
- Semantic Web and Ontologies
- Metaheuristic Optimization Algorithms Research
- Advanced Clustering Algorithms Research
- EEG and Brain-Computer Interfaces
- AI-based Problem Solving and Planning
- Cell Image Analysis Techniques
- Peer-to-Peer Network Technologies
- Mobile Agent-Based Network Management
- Cognitive Computing and Networks
- Digital Marketing and Social Media
- Human Mobility and Location-Based Analysis
- Expert finding and Q&A systems
- Remote-Sensing Image Classification
- Simulation and Modeling Applications
Shanghai Jinyuan Senior High School
2025
Harbin Engineering University
2014-2024
Wake Forest University
2024
Beijing Institute of Technology
2024
Xi'an Jiaotong University
2019-2023
Beijing University of Technology
2020-2023
Beijing University of Civil Engineering and Architecture
2021
Jiangxi University of Finance and Economics
2020
Huazhong University of Science and Technology
2011
Harbin University
2006-2009
In this paper, deep learning technology, along with a Gated Recurrent Unit (GRU) combined an attention mechanism, is used to enhance the recognition ability and risk assessment accuracy of abnormal trading behavior in financial markets. The GRU effectively solves problem gradient vanishing traditional recurrent neural networks through its unique gated structure, allowing model learn more stable effective feature representations long sequence data. On basis, contextual (CA) module mechanism...
Artificially making clinical decisions for patients with multi-morbidity has long been considered a thorny problem due to the complexity of disease. Drug recommendations can assist doctors in automatically providing effective and safe drug combinations conducive treatment reducing adverse reactions. However, existing recommendation works ignored two critical information. (i) Different types medical information their interrelationships patient's visit history be used construct comprehensive...
In recent years, information-theoretic generalization bounds have emerged as a promising approach for analyzing the capabilities of meta-learning algorithms. However, existing results are confined to two-step bounds, failing provide sharper characterization meta-generalization gap that simultaneously accounts environment-level and task-level dependencies. This paper addresses this fundamental limitation by establishing novel single-step meta-learning. Our exhibit substantial advantages over...
Abstract Feruloyl esterases (FEs, EC 3.1.1.73) play a crucial role in biological synthesis and metabolism. However, the identification of versatile FEs, capable catalyzing wide range substrates, remains challenge. In this study, we obtained 2085 FE sequences from BRENDA database initiated with an enzyme similarity network analysis, revealing three main clusters (1–3). Notably, both cluster 1 3 included characterized which exhibited significant differences sequence length. Subsequent...
At present, reliable and precise ship detection in high-resolution optical remote sensing images affected by wave clutter, thin clouds, islands under complex sea conditions is still challenging. the same time, object algorithms satellite are challenged color, aspect ratio, background, angle variability. Even results obtained based on latest convolutional neural network (CNN) method not satisfactory. In order to obtain more accurate results, this paper proposes a image brainlike visual...
Knowledge graphs as external information has become one of the mainstream directions current recommendation systems. Various knowledge-graph-representation methods have been proposed to promote development knowledge in related fields. Knowledge-graph-embedding can learn entity and complex relationships between entities graphs. Furthermore, recently graph neural networks higher-order representations Therefore, complete presentation enriches item alleviates cold start process too-sparse data....
As an important factor for improving recommendations, time information has been introduced to model users' dynamic preferences in many papers. However, the sequence of behaviour is rarely studied recommender systems. Due unique behavior evolution patterns and personalized interest transitions among items, similarity sequential dimension should be further distinguish interests. In this paper, we propose a new collaborative filtering recommendation method based on sequences (IS) that rank...
Recently, contrastive learning for sequential recommendation has demonstrated its powerful ability to learn high-quality user representations. However, constructing augmented samples in the time domain poses challenges due various reasons, such as fast-evolving trends, interest shifts, and system factors. Furthermore, F-principle indicates that deep preferentially fits low-frequency part, resulting poor performance on high-frequency tasks. The complexity of series preference limit utility...
Telemarketing has an important application in commercial promotion, and blind product recommendation a high failure rate. However, to potential users can effectively reduce marketing costs increase revenue. In this paper, 41,188 data on telemarketing from Portuguese banking institution are selected with the classification objective of predicting whether customer will subscribe time deposit account or not. The paper first preprocesses fill missing data. Secondly, describes four models used...
Consolidating semantically rich annotation on digital histopathological images known as whole-slide requires a software capable of handling such type biomedical data with support for procedures which align existing pathological protocols. Demands large-scale annotated datasets are the raise since they needed developments artificial intelligence techniques to promote automated diagnosis, mass screening, phenotype-genotype association study, etc. This paper presents an open platform efficient...
Link prediction is an important content in the related fields of social networks. Nowadays, link algorithm based on node similarity research hotspot. Node usually calculated by common neighbours, however, impact each neighbour different, and how to distinguish different roles neighbours needs be further studied. The paper proposes concept strength redefines calculation method similarity. new takes into account between nodes predicted nodes, better reflect effect node. carried out experiments...
The classification of aerial scenes has been extensively studied as the basic work remote sensing image processing and interpretation. However, performance scene based on deep neural networks is limited by number labeled samples. In order to alleviate demand for massive samples, various methods have proposed apply semi-supervised learning train classifier using unlabeled considering complex contextual relationship huge spatial differences, existing bring different degrees incorrectly samples...
Abstract The study of histopathological phenotypes is vital for cancer research and medicine as it links molecular mechanisms to disease prognosis. It typically involves integration heterogenous features in whole-slide images (WSI) objectively characterize a phenotype. However, the large-scale implementation phenotype characterization has been hindered by fragmentation features, resulting from lack standardized format controlled vocabulary structured unambiguous representation semantics...
Product tolerance is one of the key factors which can determine good or bad performance mechanical products. Its size not only affects manufacturing and assembly process, but also product features [1]. Thus optimization design gets more attention. In this paper, an improved physical programming method used to make mathematical modeling for allocation problem dimensional chain, PSO algorithm improve solving ability. And effective solution designed.