Associate Professor Lina Yao1
1UNSW
Generally, the success of AI, machine learning, especially deep learning, is relying on large scale datasets. However, it is not practical for us to access sufficient, high quality training data due to some concerns including security, privacy, safety. Therefore, the most common situation we are facing is how to develop the learning systems to efficiently learn and generalise from very few samples. In this talk, I will describe our recent work on few shot machine learning and discuss some aspects of new research opportunities and challenges.
Biography:
Associate Professor Lina Yao has made substantial research contributions towards learning contextualised actionable intelligence from sparse, diverse and dynamic multi-source data through developing innovative and effective methodologies and algorithms. Her research in generalisable and explainable data-efficient machine learning has laid a foundation that enables advanced personalised and trustworthy AI-powered human-data cooperation and applies directly to Healthcare Informatics, Cyber Security, Transportation, Defence, Industry 4.0 and FinTech.