Skip to main content

Research Repository

Advanced Search

Achieving goals using reward shaping and curriculum learning

Studley, Matthew; hansen, mark; anca, mihai; thomas, johnathan; pedamonti, dabal

Achieving goals using reward shaping and curriculum learning Thumbnail


Authors

Profile Image

Dr Matthew Studley Matthew2.Studley@uwe.ac.uk
Professor of Ethics & Technology/School Director (Research & Enterprise)

Mark Hansen Mark.Hansen@uwe.ac.uk
Professor of Machine Vision and Machine Learning

mihai anca

johnathan thomas

dabal pedamonti



Abstract

Real-time control for robotics is a popular research area in the reinforcement learning community. Through the use of techniques such as reward shaping, researchers have managed to train online agents across a multitude of domains. Despite these advances, solving goal oriented tasks still requires complex architectural changes or hard constraints to be placed on the problem. In this article, we solve the problem of stacking multiple cubes by combining curriculum learning, reward shaping, and a high number of efficiently parallelized environments. We introduce two curriculum learning settings that allow us to separate the complex task into sequential sub-goals, hence enabling the learning of a problem that may otherwise be too difficult. We focus on discussing the challenges encountered while implementing them in a goal-conditioned environment. Finally, we extend the best configuration identified on a higher complexity environment with differently shaped objects.

Citation

Studley, M., hansen, M., anca, M., thomas, J., & pedamonti, D. (2023, November). Achieving goals using reward shaping and curriculum learning. Paper presented at Future Technologies Conference, San Francisco

Presentation Conference Type Conference Paper (unpublished)
Conference Name Future Technologies Conference
Conference Location San Francisco
Start Date Nov 2, 2023
End Date Nov 3, 2023
Deposit Date May 16, 2023
Publicly Available Date Mar 29, 2024
Series Title Lecture Notes in Networks and Systems
Keywords reinforcement learning, curriculum learning, reward shaping, robotics
Public URL https://uwe-repository.worktribe.com/output/10792709

Files







You might also like



Downloadable Citations