Work

TEA: Trajectory Encoding Augmentation for Robust and Transferable Policies in Offline Reinforcement Learning

Year: 2025

Type: conference-paper

Abstract: In this paper, we investigate offline reinforcement learning (RL) with the goal of training a single robust policy that generalizes effectively across environments with unseen dynamics.We propose a novel approach, Trajectory Encoding Augmentation (TEA), which extends the state space by integrating latent representations of environmental dynamics obtained from sequence encoders, such as autoencoders.Our findings show that incorporating these encodings with TEA improves the transferability of a single policy to novel environments with new dynamics, surpassing methods that rely solely on unmodified states.These results indicate that TEA captures critical, environment-specific characteristics, enabling RL agents to generalize effectively across dynamic conditions.

Source: European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN)

DOI: 10.14428/esann/2025.es2025-114

Authors: Batıkan Bora Ormancı, Phillip Swazinna, Steffen Udluft, Thomas Runkler

Institutions: Siemens (Germany), Technical University of Munich

Language: English

Topic: Reinforcement Learning in Robotics

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: bronze

Funder: Bundesministerium für Bildung und Forschung

OpenAlex ID: W4409479689

FWCI: 1.009

Cites:

Cited by: 1

Related to: 10