TU Darmstadt / ULB / TUbiblio

Inverse Reinforcement Learning via Nonparametric Spatio-Temporal Subgoal Modeling

Šošić, A. and Rueckert, E. and Peters, J. and Zoubir, A. M. and Koeppl, H. :
Inverse Reinforcement Learning via Nonparametric Spatio-Temporal Subgoal Modeling.
[Online-Edition: http://www.jmlr.org/papers/volume19/18-113/18-113.pdf]
In: Journal of Machine Learning Research, 19 (69) pp. 1-45.
[Article] , (2018)

Official URL: http://www.jmlr.org/papers/volume19/18-113/18-113.pdf

Abstract

Advances in the field of inverse reinforcement learning (IRL) have led to sophisticated inference frameworks that relax the original modeling assumption of observing an agent behavior that reflects only a single intention. Instead of learning a global behavioral model, recent IRL methods divide the demonstration data into parts, to account for the fact that different trajectories may correspond to different intentions, e.g., because they were generated by different domain experts. In this work, we go one step further: using the intuitive concept of subgoals, we build upon the premise that even a single trajectory can be explained more efficiently locally within a certain context than globally, enabling a more compact representation of the observed behavior. Based on this assumption, we build an implicit intentional model of the agent's goals to forecast its behavior in unobserved situations. The result is an integrated Bayesian prediction framework that significantly outperforms existing IRL solutions and provides smooth policy estimates consistent with the expert's plan. Most notably, our framework naturally handles situations where the intentions of the agent change over time and classical IRL algorithms fail. In addition, due to its probabilistic nature, the model can be straightforwardly applied in active learning scenarios to guide the demonstration process of the expert.

Item Type: Article
Erschienen: 2018
Creators: Šošić, A. and Rueckert, E. and Peters, J. and Zoubir, A. M. and Koeppl, H.
Title: Inverse Reinforcement Learning via Nonparametric Spatio-Temporal Subgoal Modeling
Language: English
Abstract:

Advances in the field of inverse reinforcement learning (IRL) have led to sophisticated inference frameworks that relax the original modeling assumption of observing an agent behavior that reflects only a single intention. Instead of learning a global behavioral model, recent IRL methods divide the demonstration data into parts, to account for the fact that different trajectories may correspond to different intentions, e.g., because they were generated by different domain experts. In this work, we go one step further: using the intuitive concept of subgoals, we build upon the premise that even a single trajectory can be explained more efficiently locally within a certain context than globally, enabling a more compact representation of the observed behavior. Based on this assumption, we build an implicit intentional model of the agent's goals to forecast its behavior in unobserved situations. The result is an integrated Bayesian prediction framework that significantly outperforms existing IRL solutions and provides smooth policy estimates consistent with the expert's plan. Most notably, our framework naturally handles situations where the intentions of the agent change over time and classical IRL algorithms fail. In addition, due to its probabilistic nature, the model can be straightforwardly applied in active learning scenarios to guide the demonstration process of the expert.

Journal or Publication Title: Journal of Machine Learning Research
Volume: 19
Number: 69
Divisions: 18 Department of Electrical Engineering and Information Technology
18 Department of Electrical Engineering and Information Technology > Institute for Telecommunications > Bioinspired Communication Systems
18 Department of Electrical Engineering and Information Technology > Institute for Telecommunications
18 Department of Electrical Engineering and Information Technology > Institute for Telecommunications > Signal Processing
Date Deposited: 28 Feb 2018 12:31
Official URL: http://www.jmlr.org/papers/volume19/18-113/18-113.pdf
Export:

Optionen (nur für Redakteure)

View Item View Item