Wei Liuمشاهده پروفایل
دانشیار
Wei Liu is an Associate Professor in Machine Learning and Director of the Future Intelligence Research Lab at the University of Technology Sydney's School of Computer Science. He holds a PhD in Machine Learning from the University of Sydney and maintains active roles as a senior IEEE member and area chair for top AI conferences including KDD, AAAI, and ICDM. Education: PhD in Machine Learning, University of Sydney His research focuses on adversarial machine learning, generative AI, cybersecurity, and multimodal learning, with particular emphasis on AI security, robustness of algorithms, and model fairness. Liu's work addresses critical challenges in developing next-generation AI systems that can withstand cyber attacks while maintaining performance with multi-modal data and balanced outcomes despite data imbalances. Analysis of his recent publications reveals a strong trend toward securing large language models against novel attack vectors while advancing multimodal learning techniques. His work spans both theoretical contributions in adversarial frameworks and practical applications in cybersecurity, transportation, and industrial systems. Scientific Awards: 3 Best Paper Awards Most Influential Paper Award at PAKDD Nominee for NSW Premier's Prizes for Early Career Researcher (2017) Liu actively supervises numerous PhD students working on adversarial attacks, robust AI models, and agricultural applications. He has secured substantial funding including ARC Discovery Projects, government grants, and industry partnerships with organizations including Agriwebb, CSIRO Data61, and AVEVA. His Future Intelligence Research Lab specifically targets three emerging challenges: AI security against cyber attacks, robustness with multi-modal data, and model fairness with imbalanced datasets. The Future Intelligence Research Lab produces next-generation AI algorithms addressing AI security vulnerabilities, multi-modal robustness challenges, and fairness issues in real-world deployment scenarios, with multiple representative papers demonstrating practical applications in each domain.









