A1258
Title: Data-driven knowledge transfer in batch Q* learning
Authors: Wenbo Jing - City University of Hong Kong (Hong Kong) [presenting]
Xi Chen - New York University (United States)
Elynn Chen - New York University (United States)
Abstract: In data-driven decision-making across marketing, healthcare, and education, leveraging large datasets from existing ventures is crucial for navigating high-dimensional feature spaces and addressing data scarcity in new ventures. Knowledge transfer in dynamic decision-making is investigated by focusing on batch stationary environments and formally defining task discrepancies through the framework of Markov decision processes (MDPs). The transfer fitted q-iteration algorithm with general function approximation is proposed, which enables direct estimation of the optimal action-state function Q* using both target and source data. Under sieve approximation, the relationship between statistical performance and the MDP task discrepancy is established, highlighting the influence of source and target sample sizes and task discrepancy on the effectiveness of knowledge transfer. Theoretical and empirical results demonstrate that the final learning error of the function is significantly reduced compared to the single-task learning rate.
