A5111513990- Stock Market Forecasting Methods
- Cooperative Communication and Network Coding
- Microbial metabolism and enzyme function
- Per- and polyfluoroalkyl substances research
- Traffic Prediction and Management Techniques
- Medical and Biological Ozone Research
- Peer-to-Peer Network Technologies
- Emotion and Mood Recognition
- Advanced Malware Detection Techniques
- Sentiment Analysis and Opinion Mining
- Advanced Computational Techniques and Applications
- Network Traffic and Congestion Control
- Remote Sensing and Land Use
- Text and Document Classification Technologies
- Remote-Sensing Image Classification
- Network Security and Intrusion Detection
- Energy Load and Power Forecasting
- Metaheuristic Optimization Algorithms Research
- Image and Video Quality Assessment
- Software-Defined Networks and 5G
UNSW Sydney
2025
Tsinghua University
2003-2024
Columbia University
2024
University of Chinese Academy of Sciences
2023-2024
National Center for Nanoscience and Technology
2024
Chinese Academy of Sciences
2024
Chinese Academy of Medical Sciences & Peking Union Medical College
2024
Academy of Medical Sciences
2024
The first Multimodal Emotion Recognition Challenge (MER 2023)1 was successfully held at ACM Multimedia. challenge focuses on system robustness and consists of three distinct tracks: (1) MER-MULTI, where participants are required to recognize both discrete dimensional emotions; (2) MER-NOISE, in which noise is added test videos for modality evaluation; (3) MER-SEMI, provides a large amount unlabeled samples semi-supervised learning. In this paper, we introduce the motivation behind challenge,...
Accurate load forecasting is crucial for efficient management, planning, and operation of modern power systems, especially in an era increasing electrical demand driven by technological advancements environmental changes. This paper explores the development hybrid Transformer-based models to improve accuracy capturing short-term fluctuations long-term dependencies energy demand. Three were developed, RNR-Transformer, GRU-Transformer, LSTM-Transformer, compared with traditional such as...
The escalating focus on data privacy poses significant challenges for collaborative neural network training, where ownership and model training/deployment responsibilities reside with distinct entities.Our community has made substantial contributions to addressing this challenge, proposing various approaches such as federated learning (FL) privacy-preserving machine based cryptographic constructs like homomorphic encryption (HE) secure multiparty computation (MPC).However, FL completely...
Perfluoroalkyl and polyfluoroalkyl substances (PFAS) are widely used in industrial household products, raising serious concerns due to their environmental persistence mobility. Epidemiological studies have reported potential carcinogenic risks of PFAS based on widespread occurrence population exposure. In this study, we observed that perfluorooctanoic acid (PFOA), a common PFAS, functions as mechanical regulator lung cancer cells. PFOA exposure reduces cell stiffness, thereby decreasing...
The escalating focus on data privacy poses significant challenges for collaborative neural network training, where ownership and model training/deployment responsibilities reside with distinct entities. Our community has made substantial contributions to addressing this challenge, proposing various approaches such as federated learning (FL) privacy-preserving machine based cryptographic constructs like homomorphic encryption (HE) secure multiparty computation (MPC). However, FL completely...
This paper focuses on the application and optimization of LSTM model in financial risk prediction. The study starts with an overview architecture algorithm foundation LSTM, then details training process hyperparameter tuning strategy, adjusts network parameters through experiments to improve performance. Comparative show that optimized shows significant advantages AUC index compared random forest, BP neural XGBoost, which verifies its efficiency practicability field prediction, especially...
Undesirable information such as the computer virus, route flap, always origins in one area of network and spreads to others, thus causes far-reaching instability network. Upgrading nodes with new software and/or hardware will help suppress instability. This paper presents a hierarchical solution establish an efficient upgrade plan, determining which should be upgraded firstly. The easy-spread undesirable is analyzed. Different types are distinguished according their abilities deal...