Scientific Advisory Board of the Program
- John Aston, University of Cambridge, UK
- Tony Cai, University of Pennsylvania, USA
- Aurore Delaigle, University of Melbourne, Australia
- Jin-Chuan Duan, National University of Singapore, Singapore
- Gareth James, Emory University, USA
- Steve Marron, University of North Carolina at Chapel Hill, USA
- Hans-Georg Müller, University of California at Davis, USA
- Yu Shyr, Vanderbilt University Medical Center, USA
- Jane-Ling Wang, University of California at Davis, USA
- Fang Yao, Peking University, China
- Ming Yuan, Columbia University, USA
Functional data nowadays are commonly observed and play an ever-growing role in financial, medical, healthcare, and other scientific domains. This makes FDA an essential field particularly in the era of AI, and simultaneously creates enormous opportunities of potential exciting applications to various fields undergoing fast development, including but not limited to precision medicine and financial technology. We propose a one-month thematic program, bringing together worldwide renowned scientists to share the latest research findings in theoretical and applied FDA, to discuss the future directions of FDA, and to forge new research collaborations with the potential for significant societal impact.
In particular, we are discussing the open challenging questions in FDA raised in the era of artificial intelligence (AI) that are driven by certain properties and complex structures possessed in modern functional data:
- High-dimensional FDA, e.g., for functional data that are sampled from different locations in a large-scale network.
- Non-Euclidean FDA, e.g., for functional data taking values in a non-Euclidean space, such as a topological or metric space that does not have a vector or inner product structure.
- FDA under non-statistical constraints such as privacy protection and budgets on computational resources.
Addressing the above questions, which will bring significant contributions and insights to sciences, motivates the development of innovative methodologies and theories of advanced FDA, calls for new algorithms that are computationally efficient, and provokes the interactions between FDA and other fields of statistical learning, for instance,
- FDA and transfer learning, aiming to improve performance of FDA procedures on a learning task by transferring knowledge that is already learned from other related tasks and data into the new task;
- Interplay between FDA and machine learning, primarily targeting at leveraging the power of modern machine learning algorithms for better modelling non-linearity in FDA.
In turn, the era of artificial intelligence (AI) creates enormous opportunities for exciting applications and solutions using disruptive innovations developed in FDA in various fields, including but not limited to machine learning, image analysis, bioscience, precision medicine, and fintech.
This one-month-long program will invite world-leading experts in the areas of FDA, high-dimensional data analysis, non-Euclidean data analysis, and machine learning, and encourage them to bring insights into the interactions between these and other fields, in e.g. data-oriented precision medicine and FinTech, and will provide valuable advice on the future development of FDA. The program will provide an ideal platform for the local and international mathematicians, statisticians and data scientists to exchange ideas and prompt development of FDA for the challenges it will need to face in the era of big data and AI. The program aims to help arouse local researchers’ interest in FDA, especially via presentations and discussions that focus on connections between FDA and other domains such as machine learning. It is to further enhance Singapore’s capacity in FDA in preparation for challenges and opportunities that are related to healthcare, medical sciences, financial technology, among many others, and eventually wellness of society.