Corpus ID: 211066514

Provably efficient reconstruction of policy networks

@article{Mazoure2020ProvablyER,
  title={Provably efficient reconstruction of policy networks},
  author={Bogdan Mazoure and Thang Doan and T. Li and V. Makarenkov and Joelle Pineau and Doina Precup and Guillaume Rabusseau},
  journal={ArXiv},
  year={2020},
  volume={abs/2002.02863}
}
  • Bogdan Mazoure, Thang Doan, +4 authors Guillaume Rabusseau
  • Published 2020
  • Mathematics, Computer Science
  • ArXiv
  • Recent research has shown that learning poli-cies parametrized by large neural networks can achieve significant success on challenging reinforcement learning problems. However, when memory is limited, it is not always possible to store such models exactly for inference, and com-pressing the policy into a compact representation might be necessary. We propose a general framework for policy representation, which reduces this problem to finding a low-dimensional embedding of a given density… CONTINUE READING

    Figures and Topics from this paper.

    Deep Reinforcement and InfoMax Learning
    3

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 42 REFERENCES
    The role of factor XI in coagulation.
    30
    D1 and D5-brane giant gravitons on AdS3 × S3 × S3 × S1
    8
    intake: Variation and relationship to other food components [Abstract
    • 1997
    Calculation of NMR and EPR parameters : theory and applications
    591
    Clustering and Classification
    514