The goal of all-inclusive finance is to bring reliable and high quality financial service to everyone everywhere, rich or poor. Managing large scale multi-agent interactions lies in the very nature of the all-inclusive finance, and many emerging problems in this new space involve sophisticated cooperation and competition between a diverse range of entities, such as people, small businesses, online platforms, financial units and even fraudsters. For instance,
Machine learning techniques, such as multi-agent reinforcement learning, algorithmic game theory, generative adversarial learning, imitation learning, graph neural networks, construction and reasoning over knowledge graph, building interpretable and fair models, are playing increasingly important roles in addressing these all-inclusive finance problems. In this talk, I will give three concrete examples of our recent work along this line: 1. Generative adversarial imitation learning for reinforcement learning based recommendation system; 2. Value Propagation for Decentralized Networked Deep Multi-agent Reinforcement Learning; 3. Double neural counterfactual regret minimization for two player zero sum game with application to Texas Hold’em poker game.
Dr. Yuan (Alan) Qi is Vice President and Chief AI Scientist of Ant Financial Services Group, and lead of Alibaba DAMO Academy Financial Intelligence. Before joining Alibaba and Ant Financial, he obtained his PhD from MIT and tenured associate professorship in Computer Science and Statistics from Purdue University.He previously served as associate editor of Journal of Machine Learning Research and area chair of International Conference on Machine Learning. He received Microsoft's Newton Breakthrough Research award in 2008 and USA NSF Career award in 2011. At Ant Financial, he leads the AI department to build AI tools and solutions to address various financial problems such as microlending, risk control, insurance claiming process, marketing and customer service, empowering both internal and external business partners.