A Broad-persistent Advising Approach for Deep Interactive Reinforcement Learning in Robotic Environments

TL;DR

This paper introduces Broad-persistent Advising (BPA), a novel approach in Deep Interactive Reinforcement Learning that retains and reuses advice to improve learning efficiency in robotic tasks.

Contribution

BPA extends existing DeepIRL by enabling the reuse of advice across similar states, enhancing learning speed and reducing trainer interactions in robotic environments.

Findings

BPA improves agent performance in robotic tasks.

BPA reduces the number of trainer interactions.

BPA demonstrates effectiveness in continuous robotic scenarios.

Abstract

Deep Reinforcement Learning (DeepRL) methods have been widely used in robotics to learn about the environment and acquire behaviors autonomously. Deep Interactive Reinforcement Learning (DeepIRL) includes interactive feedback from an external trainer or expert giving advice to help learners choosing actions to speed up the learning process. However, current research has been limited to interactions that offer actionable advice to only the current state of the agent. Additionally, the information is discarded by the agent after a single use that causes a duplicate process at the same state for a revisit. In this paper, we present Broad-persistent Advising (BPA), a broad-persistent advising approach that retains and reuses the processed information. It not only helps trainers to give more general advice relevant to similar states instead of only the current state but also allows the agent to speed up the learning process. We test the proposed approach in two continuous robotic scenarios, namely, a cart pole balancing task and a simulated robot navigation task. The obtained results show that the performance of the agent using BPA improves while keeping the number of interactions required for the trainer in comparison to the DeepIRL approach.