Corpus ID: 209324424

Measuring the Reliability of Reinforcement Learning Algorithms

@article{Chan2020MeasuringTR,
  title={Measuring the Reliability of Reinforcement Learning Algorithms},
  author={Stephanie C. Y. Chan and Sam Fishman and J. Canny and A. Balan and S. Guadarrama},
  journal={ArXiv},
  year={2020},
  volume={abs/1912.05663}
}
  • Stephanie C. Y. Chan, Sam Fishman, +2 authors S. Guadarrama
  • Published 2020
  • Mathematics, Computer Science
  • ArXiv
  • Inadequate reliability is a well-known issue for reinforcement learning (RL) algorithms. This problem has gained increasing attention in recent years, and efforts to improve it have grown substantially. To aid RL researchers and production users with the evaluation and improvement of reliability, we propose a novel set of metrics that quantitatively measure different aspects of reliability. In this work, we address variability and risk, both during training and after learning (on a fixed policy… CONTINUE READING

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 32 REFERENCES
    Deep Reinforcement Learning that Matters
    603
    Reproducibility of Benchmarked Deep Reinforcement Learning Tasks for Continuous Control
    107
    A Distributional Perspective on Reinforcement Learning
    390
    Benchmarking Deep Reinforcement Learning for Continuous Control
    831
    How Many Random Seeds? Statistical Power Analysis in Deep Reinforcement Learning Experiments
    23
    Deterministic Implementations for Reproducibility in Deep Reinforcement Learning
    13
    Proximal Policy Optimization Algorithms
    2575
    Policy Gradient Methods for Reinforcement Learning with Function Approximation
    3015
    Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
    799