Academy of Mathematics and Systems Science, CAS
Colloquia & Seminars

Speaker:	黄鑫，南方科技大学
Inviter:
Title:	A Two-level Reinforcement Learning Algorithm for Ambiguous Mean-variance Portfolio Selection Problem
Language:	Chinese
Time & Venue:	2023.09.07 14:00-15:00 思源楼S315
Abstract:	Traditional modeling on the mean-variance portfolio selection problem often assumes a full knowledge on the statistics of assets’ returns. It is, however, not always the case in the real market. This work deals with an ambiguous mean-variance portfolio selection problem with a mixture model on the returns of risky assets, where the proportions of different component distributions are assumed to be unknown to the investors but being constants. Taking into consideration the updates of proportions from future observations, we derive an investment policy with a learning feature through a two-level framework. Our approach combines the time-decomposed approach (dynamic programming) in the lower level and a scenario-decomposed approach (progressive hedging algorithm) in the upper layer. We carry out experimental studies to illustrate the execution of our policy scheme.