Provable Hierarchical Imitation Learning via EM

Zhiyu Zhang, Ioannis Paschalidis
Proceedings of The 24th International Conference on Artificial Intelligence and Statistics, PMLR 130:883-891, 2021.

Abstract

Due to recent empirical successes, the options framework for hierarchical reinforcement learning is gaining increasing popularity. Rather than learning from rewards, we consider learning an options-type hierarchical policy from expert demonstrations. Such a problem is referred to as hierarchical imitation learning. Converting this problem to parameter inference in a latent variable model, we develop convergence guarantees for the EM approach proposed by Daniel et al. (2016b). The population level algorithm is analyzed as an intermediate step, which is nontrivial due to the samples being correlated. If the expert policy can be parameterized by a variant of the options framework, then, under regularity conditions, we prove that the proposed algorithm converges with high probability to a norm ball around the true parameter. To our knowledge, this is the first performance guarantee for an hierarchical imitation learning algorithm that only observes primitive state-action pairs.

Cite this Paper

BibTeX
@InProceedings{pmlr-v130-zhang21b, title = { Provable Hierarchical Imitation Learning via EM }, author = {Zhang, Zhiyu and Paschalidis, Ioannis}, booktitle = {Proceedings of The 24th International Conference on Artificial Intelligence and Statistics}, pages = {883--891}, year = {2021}, editor = {Banerjee, Arindam and Fukumizu, Kenji}, volume = {130}, series = {Proceedings of Machine Learning Research}, month = {13--15 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v130/zhang21b/zhang21b.pdf}, url = {https://proceedings.mlr.press/v130/zhang21b.html}, abstract = { Due to recent empirical successes, the options framework for hierarchical reinforcement learning is gaining increasing popularity. Rather than learning from rewards, we consider learning an options-type hierarchical policy from expert demonstrations. Such a problem is referred to as hierarchical imitation learning. Converting this problem to parameter inference in a latent variable model, we develop convergence guarantees for the EM approach proposed by Daniel et al. (2016b). The population level algorithm is analyzed as an intermediate step, which is nontrivial due to the samples being correlated. If the expert policy can be parameterized by a variant of the options framework, then, under regularity conditions, we prove that the proposed algorithm converges with high probability to a norm ball around the true parameter. To our knowledge, this is the first performance guarantee for an hierarchical imitation learning algorithm that only observes primitive state-action pairs. } }
Endnote
%0 Conference Paper %T Provable Hierarchical Imitation Learning via EM %A Zhiyu Zhang %A Ioannis Paschalidis %B Proceedings of The 24th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2021 %E Arindam Banerjee %E Kenji Fukumizu %F pmlr-v130-zhang21b %I PMLR %P 883--891 %U https://proceedings.mlr.press/v130/zhang21b.html %V 130 %X Due to recent empirical successes, the options framework for hierarchical reinforcement learning is gaining increasing popularity. Rather than learning from rewards, we consider learning an options-type hierarchical policy from expert demonstrations. Such a problem is referred to as hierarchical imitation learning. Converting this problem to parameter inference in a latent variable model, we develop convergence guarantees for the EM approach proposed by Daniel et al. (2016b). The population level algorithm is analyzed as an intermediate step, which is nontrivial due to the samples being correlated. If the expert policy can be parameterized by a variant of the options framework, then, under regularity conditions, we prove that the proposed algorithm converges with high probability to a norm ball around the true parameter. To our knowledge, this is the first performance guarantee for an hierarchical imitation learning algorithm that only observes primitive state-action pairs.
APA
Zhang, Z. & Paschalidis, I.. (2021). Provable Hierarchical Imitation Learning via EM . Proceedings of The 24th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 130:883-891 Available from https://proceedings.mlr.press/v130/zhang21b.html.

Related Material