Sign inSign in
ConceptAI & ML

Policy

The agent's behavior function mapping states to actions.
as David Silver defines itGoogle DeepMindThe wording closest to all 11 definitions
Watch it explained

Know first

Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.

Opens up

Concepts that name this one as a prerequisite.

Lecturers

8

Each in their own terms, from the easiest level.

  • David Silver

    Google DeepMind · Introductory

    A distribution over actions given states that fully defines the behavior of an agent.
    • 46:16Markov Processes, Reward Processes & Bellman Equations
  • Pieter Abbeel

    UC Berkeley · Undergraduate

    A mapping from states to actions that dictates what an agent should do in any given state.
    • 16:20Markov Decision Processes (MDPs) Part 1/2
    • 13:17Markov Decision Processes (MDPs) Part 2/2
  • Emma Brunskill

    Stanford University · Undergraduate

    A mapping from states to actions that defines the behavior of an agent.
    • 42:17Introduction to Reinforcement Learning (CS234, 2024, Lecture 1)
  • Fei-Fei Li, Justin Johnson & Serena Yeung

    Stanford University · Undergraduate

    A function from states to actions that specifies what action to take in each state.
    • 8:08Deep Reinforcement Learning
  • Russ Tedrake

    MIT · Undergraduate

    A function that maps states to control actions.
    • 22:36Dynamic Programming I
  • Sergey Levine

    UC Berkeley · Graduate

    A distribution over actions conditional on observations or states.
    • 0:21Policy Gradient Foundations: Trajectory Distributions & High-Dimensional State Spaces
    • 0:30Imitation Learning I: Behavioral Cloning & Distributional Shift
  • Leslie Kaelbling

    MIT · Graduate

    A closed-loop controller that takes the system from one state to another.
    • 10:38The Role of Rationality in Modern Robotics
  • Max Simchowitz

    Carnegie Mellon University · Research

    A mapping from states to distributions over actions that describes what an agent does.
    • 5:44Generative Control, Action Chunking, and Moravec’s Paradox

Lectures that assume it

15

The first 12 are shown.

Connections

35

The lectures that define it or assume it, and the concepts next to it.

Concepts9

Show 3 more

Lectures26

Show 20 more

8 lecturers. Definitions are written by AI from each lecture’s transcript. A time with a play mark starts the lecture at the moment the lecturer says it.