Similarity-based transfer learning of decision policies

2020/06/12 by Zugarová, Eliška, Guy, Tatiana V.
#68T05 (Primary) #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Systems and Control (eess.SY) #electronic engineering #information engineering

paper · doi:10.48550/arxiv.2006.08768

Abstract

A problem of learning decision policy from past experience is considered. Using the Fully Probabilistic Design (FPD) formalism, we propose a new general approach for finding a stochastic policy from the past data.

Related

• Clustering Markov Decision Processes For Continual Transfer
• Relative Policy-Transition Optimization for Fast Policy Transfer
• Learning Robust Decision Policies from Observational Data
• A Taxonomy of Similarity Metrics for Markov Decision Processes
• Transfer learning with causal counterfactual reasoning in Decision Transformers