A5103027656- Hydrological Forecasting Using AI
- IoT-based Smart Home Systems
- Water Quality Monitoring Technologies
- Advanced Computational Techniques and Applications
- Privacy-Preserving Technologies in Data
- Web Applications and Data Management
- Neural Networks and Applications
- Indoor and Outdoor Localization Technologies
- IoT and Edge/Fog Computing
- Simulation and Modeling Applications
- Advanced Decision-Making Techniques
- Patient Dignity and Privacy
- Mobile Agent-Based Network Management
- Mobile Crowdsensing and Crowdsourcing
- RFID technology advancements
- IPv6, Mobility, Handover, Networks, Security
- Advanced Malware Detection Techniques
- Industrial Technology and Control Systems
- Cloud Computing and Resource Management
- Adversarial Robustness in Machine Learning
- Service-Oriented Architecture and Web Services
- Water Quality Monitoring and Analysis
- Advanced Algorithms and Applications
Hainan University
2010-2023
An accurate prediction of cage-cultured water quality is a hot topic in smart mariculture. Since the mariculturing environment always open to its surroundings, changes parameters are normally nonlinear, dynamic, changeable, and complex. However, traditional forecasting methods have lots problems, such as low accuracy, poor generalization, high time complexity. In order solve these shortcomings, novel method based on deep LSTM (long short-term memory) learning network proposed predict pH...
Properly allocation of virtual machines is important for computing infrastructures scheduling. This paper presents systemic method on machine array optimization control based artificial intelligence and matrix theory. According to request service data from users provide proper VMs roughly via intelligent pattern recognition RBFNN, the sent a multiple-targets process produce precisely, thus enable minimize cast enhance efficiency whole achieve low consumption ensure stability system....
Split Federated Learning (SFL) is the most recent distributed training scheme. Compared to Learning, SFL reduces overhead of client while achieving better privacy protection. However, attackers can still use intermediate activations reconstruct original data containing sensitive information. To defend against this reconstruction attack, we propose distance correlation loss reduce overall between input and activations, further construct an effective efficient dynamic channel pruning network...
Mobile crowd sensing (MCS) network collects scenario, environmental, and individual data within a specific range via the intelligent equipment carried by mobile users, thus providing social decision-making services. MCS is emerging as most important paradigm. However, person-centered itself carries risk of divulging users’ privacy. To address this problem, we proposed variable weight privacy-preserving algorithm secure multiparty computation. This based on utility its effectiveness...