Axioms for Rational Reinforcement Learning

@inproceedings{Sunehag2011AxiomsFR,
  title={Axioms for Rational Reinforcement Learning},
  author={Peter Sunehag and Marcus Hutter},
  booktitle={ALT},
  year={2011}
}
  • Peter Sunehag, Marcus Hutter
  • Published in ALT 2011
  • Mathematics, Computer Science
  • We provide a formal, simple and intuitive theory of rational decision making including sequential decisions that affect the environment. The theory has a geometric flavor, which makes the arguments easy to visualize and understand. Our theory is for complete decision makers, which means that they have a complete set of preferences. Our main result shows that a complete rational decision maker implicitly has a probabilistic model of the environment. We have a countable version of this result… CONTINUE READING

    Topics from this paper.

    Rationality, optimism and guarantees in general reinforcement learning
    11
    Optimistic AIXI
    8
    Artificial General Intelligence
    One Decade of Universal Artificial Intelligence
    29
    A Dual Process Theory of Optimistic Cognition
    2
    Principles of Solomonoff Induction and AIXI
    6

    References

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