Learning From Explanations Using Sentiment and Advice in RL

@article{Krening2017LearningFE,
  title={Learning From Explanations Using Sentiment and Advice in RL},
  author={Samantha Krening and Brent Harrison and Karen M. Feigh and Charles Lee Isbell and Mark O. Riedl and Andrea Lockerd Thomaz},
  journal={IEEE Transactions on Cognitive and Developmental Systems},
  year={2017},
  volume={9},
  pages={44-55}
}
In order for robots to learn from people with no machine learning expertise, robots should learn from natural human instruction. Most machine learning techniques that incorporate explanations require people to use a limited vocabulary and provide state information, even if it is not intuitive. This paper discusses a software agent that learned to play the Mario Bros. game using explanations. Our goals to improve learning from explanations were twofold: 1) to filter explanations into advice and… CONTINUE READING

Figures, Tables, Results, and Topics from this paper.

Citations

Publications citing this paper.
SHOWING 1-10 OF 22 CITATIONS

Producing Radiologist-Quality Reports for Interpretable Deep Learning.

  • 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019)
  • 2019
VIEW 1 EXCERPT
CITES METHODS

Regional Tree Regularization for Interpretability in Black Box Models

Mike Wu, Sonali Parbhoo, +5 authors Finale Doshi-Velez
  • ArXiv
  • 2019
VIEW 1 EXCERPT
CITES BACKGROUND

A Survey of Methods for Explaining Black Box Models

  • ACM Comput. Surv.
  • 2018
VIEW 1 EXCERPT
CITES BACKGROUND

References

Publications referenced by this paper.
SHOWING 1-10 OF 22 REFERENCES

Grounding English Commands to Reward Functions

  • Robotics: Science and Systems
  • 2015
VIEW 1 EXCERPT

Robot Learning from Human Teachers

  • Robot Learning from Human Teachers
  • 2014
VIEW 2 EXCERPTS