Similarity-based transfer learning of decision policies

TL;DR

This paper introduces a novel transfer learning approach for decision policies using Fully Probabilistic Design, enabling the derivation of stochastic policies from past experience.

Contribution

It presents a new general method for learning stochastic decision policies from historical data within the FPD framework.

Findings

Effective policy transfer demonstrated on benchmark tasks

Improved decision-making accuracy over traditional methods

Framework adaptable to various decision-making scenarios

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.