A5054758911- Adversarial Robustness in Machine Learning
- Privacy-Preserving Technologies in Data
- Energy Load and Power Forecasting
- Artificial Intelligence in Healthcare
- Wheat and Barley Genetics and Pathology
- Big Data and Digital Economy
- Cryptography and Data Security
- Corrosion Behavior and Inhibition
- Radiomics and Machine Learning in Medical Imaging
- Artificial Intelligence in Healthcare and Education
- Concrete Corrosion and Durability
- Internet Traffic Analysis and Secure E-voting
- Privacy, Security, and Data Protection
- Building Energy and Comfort Optimization
- Cardiovascular Health and Risk Factors
- Structural Behavior of Reinforced Concrete
- Genetic Mapping and Diversity in Plants and Animals
- Noise Effects and Management
- Smart Grid Security and Resilience
- Genetics and Plant Breeding
Shandong Jianzhu University
2025
Harbin Institute of Technology
2018-2024
Dezhou University
2016
Trustworthy artificial intelligence (TAI) has proven invaluable in curbing potential negative repercussions tied to AI applications. Within the TAI spectrum, federated learning (FL) emerges as a promising solution safeguard personal information distributed settings across multitude of practical contexts. However, realm FL is not without its challenges. Especially worrisome are adversarial attacks targeting algorithmic robustness and systemic confidentiality. Moreover, presence biases opacity...
The rapid success of Large Language Models (LLMs) has unlocked vast potential for AI applications in privacy-sensitive domains. However, the traditional centralized training LLMs poses significant challenges due to privacy concerns regarding collecting sensitive data from diverse sources. This paper offers a promising and privacy-enhancing solution LLMs: collaboratively via Federated Learning (FL) across multiple clients, eliminating need raw transmission. To this end, we present F4LLM, new...
SUMMARY Flour whiteness (FW) is an important factor in assessing flour quality and determining the end product quality. It integrated sensory indicator reflecting colour negatively correlated with protein content. In order to dissect genetic relationship between FW its five related traits at quantitative trait locus (QTL)/gene level, a recombinant inbred line population was evaluated under three environments. Quantitative loci for were analysed by unconditional conditional QTL mapping. Four...
Cardiovascular diseases (CVDs) are currently the leading cause of death worldwide, highlighting critical need for early diagnosis and treatment. Machine learning (ML) methods can help diagnose CVDs early, but their performance relies on access to substantial data with high quality. However, sensitive nature healthcare often restricts individual clinical institutions from sharing train sufficiently generalized unbiased ML models. Federated Learning (FL) is an emerging approach, which offers a...
Recent advancements in pre-trained large language models (LLMs) have significantly influenced various domains. Adapting these for specific tasks often involves fine-tuning (FT) with private, domain-specific data. However, privacy concerns keep this data undisclosed, and the computational demands deploying LLMs pose challenges resource-limited holders. This has sparked interest split learning (SL), a Model-as-a-Service (MaaS) paradigm that divides into smaller segments distributed training...
As pre-trained models, like Transformers, are increasingly deployed on cloud platforms for inference services, the privacy concerns surrounding model parameters and data becoming more acute. Current Privacy-Preserving Transformer Inference (PPTI) frameworks struggle with "impossible trinity" of privacy, efficiency, performance. For instance, Secure Multi-Party Computation (SMPC)-based solutions offer strong guarantees but come significant overhead performance trade-offs. On other hand, PPTI...
This paper presents an effective method of parameter ascertainment for the skeleton curve corroded compression‐bending members to establish its restoring force model. An assumption which considers damaged and undamaged has similar shape is introduced into fitting process parameters. Meanwhile, two‐dimensional plane section used simplify mathematical model reduce computational cost. Several sets experimental data were compared with prediction by developed in this paper, verification. The case...