FINTECH THRUST SEMINAR | From Langevin sampling algorithms to Generative AI: Applications and Challenges in Investing
Supporting the below United Nations Sustainable Development Goals:支持以下聯合國可持續發展目標：支持以下联合国可持续发展目标：
From Langevin sampling algorithms to Generative AI: Applications and Challenges in Investing
Abstract:
We will review some important progress in Langevin sampling algorithms and their success in solving challenging, non-convex optimization problems including the training of (artificial) neural networks. We will link all this with the very recent progress in foundational research for the development of (diffusion-based) generative AI. Some intriguing applications will be discussed along with their associated challenges.
Professor Sotirios Sabanis holds the personal chair in Stochastic Analysis and Algorithms at the School of Mathematics of the University of Edinburgh.
He is the Director of the Centre for Investing Innovation, a strategic partnership between global asset management company abrdn and the University of Edinburgh. The Centre is funded by a £7.5 million commitment from abrdn across a five-year period.
He collaborates with industry (mainly AI and financial services) through a number of joint projects. His research portfolio includes funding from the Alan Turing Institute (Fellowship), The Royal Society, EPSRC and Innovate UK.