reinforcement learning
play

Reinforcement Learning Philipp Koehn 16 April 2020 Philipp Koehn - PowerPoint PPT Presentation

Jun 09, 2023 •554 likes •1.06k views

Reinforcement Learning Philipp Koehn 16 April 2020 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020 Rewards 1 Agent takes actions Agent occasionally receives reward Maybe just at the end of the process,

  1. Reinforcement Learning Philipp Koehn 16 April 2020 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  2. Rewards 1 ● Agent takes actions ● Agent occasionally receives reward ● Maybe just at the end of the process, e.g., Chess: – agent has to decide on individual moves – reward only at end: win/lose ● Maybe more frequently – Scrabble: points for each word played – ping pong: any point scored – baby learning to crawl: any forward movement Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  3. Markov Decision Process 2 State Map Stochastic Movement ● States s ∈ S , actions a ∈ A ● Model T ( s,a,s ′ ) ≡ P ( s ′ ∣ s,a ) = probability that a in s leads to s ′ ● Reward function R ( s ) (or R ( s,a ) , R ( s,a,s ′ ) ) = { − 0 . 04 (small penalty) for nonterminal states ± 1 for terminal states Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  4. Agent Designs 3 ● Utility based agent – needs model of environment – learns utility function on states – selects action that maximize expected outcome utility ● Q-learning – learns action-utility function ( Q ( s,a ) function) – does not need to model outcomes of actions – function provides expected utility of taken a given action at a given step ● Reflex agent – learns policy that maps states to actions Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  5. 4 passive reinforcement learning Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  6. Setup 5 Reward Function State Map Stochastic Movement ⎧ ⎪ ⎪ ⎪ +1 for goal ⎨ ⎪ R(s) = –1 ⎪ for pit ⎪ ⎩ –0.04 for other ● We know which state we are in (= partially observable environment) ● We know which actions we can take ● But only after taking an action → new state becomes known → reward becomes known Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  7. Passive Reinforcement Learning 6 ● Given a policy ● Task: compute utility of policy ● We will extend this later to active reinforcement learning ( ⇒ policy needs to be learned) Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  8. Sampling 7 -0.04 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  9. Sampling 8 -0.04 -0.04 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  10. Sampling 9 -0.04 -0.04 -0.04 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  11. Sampling 10 -0.04 -0.04 -0.04 -0.04 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  12. Sampling 11 -0.04 -0.04 -0.04 -0.04 -0.04 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  13. Sampling 12 -0.04 -0.04 -0.04 -0.04 -0.04 -0.04 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  14. Sampling 13 -0.04 -0.04 -0.04 +1 -0.04 -0.04 -0.04 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  15. Sampling 14 0.92 0.96 1.00 0.80 0.88 0.84 0.76 0.72 ● Sample of reward to go Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  16. Sampling 15 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  17. Sampling 16 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  18. Utility of Policy 17 ● Definition of utility U of the policy π for state s ∞ U π ( s ) = E [ γ t R ( S t )] ∑ t = 0 ● Start at state S 0 = s ● Reward for state is R ( s ) ● Discount factor γ (we use γ = 1 in our examples) Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  19. Direct Utility Estimation 18 ● Learning from the samples ● Reward to go: 0.92 0.96 1.00 0.80 – (1,1) one sample: 0.72 0.88 0.84 – (1,2) two samples: 0.76, 0.84 – (1,3) two samples: 0.80, 0.88 0.76 ● Reward to go 0.72 will converge to utility of state ● But very slowly — can we do better? Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  20. Bellman Equation 19 ● Direct utility estimation ignores dependency between states ● Given by Bellman equation U π ( s ) = R ( s ) + γ ∑ P ( s ′ ∣ s,π ( s )) U π ( s ′ ) s ′ ( γ = reward decay) ● Use of this known dependence can speed up learning ● Requires learning of transition probabilities P ( s ′ ∣ s,π ( s )) Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  21. Adaptive Dynamic Programming 20 Need to learn: ● State rewards R ( s ) – whenever a state is visited, record award (deterministic) ● Outcome of action π ( s ) at state s according to policy π – collect statistic count ( s,s ′ ) that s ′ is reached from s – estimate probability distribution count ( s,s ′ ) P ( s ′ ∣ s,π ( s )) = ∑ s ′′ count ( s,s ′′ ) ⇒ Ingredients for policy evaluation algorithm Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  22. Adaptive Dynamic Programming 21 Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  23. Learning Curve 22 ● Major change at 78 th trial: first time terminated in –1 state at (4,2) Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  24. Temporal Difference Learning 23 ● Idea: do not model P ( s ′ ∣ s,π ( s )) , directly adjust utilities U ( s ) for all visited states ● Estimate of current utility: U π ( s ) ● Estimate of utility after action: R ( s ) + γU π ( s ′ ) ● Adjust utility of current state U π ( s ) if they differ ∆ U π ( s ) = α ( R ( s ) + γU π ( s ′ ) − U π ( s )) ( α = learning rate) ● Learning rate may decrease when state has been visited often Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  25. Learning Curve 24 ● Noisier, converging more slowly Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  26. Comparison 25 ● Both eventually converge to correct values ● Adaptive dynamic programming (ADP) faster than temporal difference learning (TD) – both make adjustments to make successors agree – but: ADP adjusts all possible successors, TD only observed successor ● ADP computationally more expensive due to policy evaluation algorithm Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  27. 26 active reinforcement learning Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  28. Active Reinforcement Learning 27 ● Previously: passive agent follows prescribed policy ● Now: active agent decides which action to take – following optimal policy (as currently viewed) – exploration ● Goal: optimize rewards for a given time frame Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  29. Greedy Agent 28 1. Start with initial policy 2. Compute utilities (using ADP) 3. Optimize policy 4. Go to Step 2 ● This very seldom converges to global optimal policy Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  30. Learning Curve 29 ● Greedy agent stuck in local optimum Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  31. Bandit Problems 30 ● Bandit: slot machine Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  32. Bandit Problems 31 ● Bandit: slot machine ● N-armed bandit: n levers ● Each has different probability distribution over payoffs ● Spend coin on – presume optimal payoff – exploration (new lever) ● If independent – Gittins index : formula for solution – uses payoff / number of times used Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  33. Greedy in the Limit of Infinite Exploration 32 ● Explore any action in any state unbounded number of times ● Eventually has to become greedy – carry out optimal policy ⇒ maximize reward ● Simple strategy – with probability p ( 1 / t ) take random action – initially ( t small) focus on exploration – later ( t big) focus on optimal policy Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  34. Extension of Adaptive Dynamic Programming 33 ● Previous definition of utility calculation U ( s ) ← R ( s ) + γ max a ∑ P ( s ′ ∣ s,a ) U ( s ′ ) s ′ ● New utility calculation U + ( s ) ← R ( s ) + γ max a f (∑ P ( s ′ ∣ s,a ) U + ( s ′ ) ,N ( s,a )) s ′ ● One possible definition of f ( u,n ) f ( u,n ) = { R + if n < N c u otherwise R + is optimistic estimate, best possible award in any state Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

  35. Learning Curve 34 ● Performance of exploratory ADP agent ● Parameter settings R + = 2 and N e = 5 ● Fairly quick convergence to optimal policy Philipp Koehn Artificial Intelligence: Reinforcement Learning 16 April 2020

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

RL Overview of topics About Reinforcement Learning The Reinforcement Learning Problem

RL Overview of topics About Reinforcement Learning The Reinforcement Learning Problem

Introduction to Reinforcement Learning RL Overview of topics About Reinforcement Learning The Reinforcement Learning Problem Inside an RL agent Temporal difference learning Many faces of Reinforcement Learning What is

551 views • 35 slides
Reinforcement Learning Reinforcement Learning Reinforcement Learning in a nutshell g Imagine

Reinforcement Learning Reinforcement Learning Reinforcement Learning in a nutshell g Imagine

Reinforcement Learning Reinforcement Learning Reinforcement Learning in a nutshell g Imagine playing a new game whose rules you dont know; after a hundred or so moves your don t know; after a hundred or so moves, your opponent announces, You

511 views • 30 slides
Reinforcement Learning You can think of supervised learning as the teacher providing answers

Reinforcement Learning You can think of supervised learning as the teacher providing answers

Reinforcement Learning You can think of supervised learning as the teacher providing answers (the class labels) Reinforcement learning: In reinforcement learning, the agent learns Markov Decision Processes based on a punishment/reward

89 views • 6 slides
Examples of Reinforcement Learning Robocup Soccer Teams Stone &amp; Veloso, Reidmiller et al.

Examples of Reinforcement Learning Robocup Soccer Teams Stone &amp; Veloso, Reidmiller et al.

Reinforcement Learning Slides by Rich Sutton Mods by Dan Lizotte Refer to Reinforcement Learning: An Introduction by Sutton and Barto Alpaydin Chapter 16 Up until now we have been Supervised Learning Classifying, mostly

891 views • 31 slides
CS885 Reinforcement Learning Module 2: June 6, 2020 Maximum Entropy Reinforcement Learning

CS885 Reinforcement Learning Module 2: June 6, 2020 Maximum Entropy Reinforcement Learning

CS885 Reinforcement Learning Module 2: June 6, 2020 Maximum Entropy Reinforcement Learning Haarnoja, Tang et al. (2017) Reinforcement Learning with Deep Energy Based Policies, ICML . Haarnoja, Zhou et al. (2018) Soft Actor-Critic: Off-Policy

682 views • 24 slides
Reinforcement Learning Reinforcement Learning Now that you know a little about Optimal Control

Reinforcement Learning Reinforcement Learning Now that you know a little about Optimal Control

Reinforcement Learning Reinforcement Learning Now that you know a little about Optimal Control Theory, you actually have some knowledge in RL. RL shares the overall goal with OCT: solving for a control policy such that the cumulative cost

610 views • 22 slides
Reinforcement Learning CS 4100: Artificial Intelligence Reinforcement Learning II Still

Reinforcement Learning CS 4100: Artificial Intelligence Reinforcement Learning II Still

Reinforcement Learning CS 4100: Artificial Intelligence Reinforcement Learning II Still assume a Marko kov decision process (MDP): A se set o of st states s s S A se set o of a actions ( s (per st state) A A mo model

96 views • 5 slides
R i f R i f Reinforcement Learning III Reinforcement Learning III t L t L i i III III Dec

R i f R i f Reinforcement Learning III Reinforcement Learning III t L t L i i III III Dec

R i f R i f Reinforcement Learning III Reinforcement Learning III t L t L i i III III Dec 03 2008 1 Large State Spaces h When a problem has a large state space we can not longer represent the U or Q functions as explicit tables explicit

436 views • 18 slides
Reinforcement Learning UMaine COS 470/570 Introduction to AI Why reinforcement learning?

Reinforcement Learning UMaine COS 470/570 Introduction to AI Why reinforcement learning?

Reinforcement Learning&lt;br/&gt;&lt;br/&gt; 4/25/19, 8*06 PM Reinforcement Learning&lt;br/&gt;&lt;br/&gt; 4/25/19, 8*06 PM Reinforcement Learning UMaine COS 470/570 Introduction to AI Why reinforcement learning? Spring 2019 Created:

371 views • 15 slides
7. Motor Control and Reinforcement Learning Outline A. Action Selection and Reinforcement B.

7. Motor Control and Reinforcement Learning Outline A. Action Selection and Reinforcement B.

7. Motor Control and Reinforcement Learning Outline A. Action Selection and Reinforcement B. Temporal Difference Reinforcement Learning C. PVLV Model D. Cerebellum and Error-driven Learning 2/23/18 COSC 494/594 CCN 2 Sensory-Motor Loop

790 views • 56 slides
Deep Reinforcement Learning [Mastering the Game of Go with Deep Reinforcement Learning and Tree

Deep Reinforcement Learning [Mastering the Game of Go with Deep Reinforcement Learning and Tree

Deep Reinforcement Learning [Mastering the Game of Go with Deep Reinforcement Learning and Tree Search, Nature 2016] CS 486/686 University of Waterloo Lecture 21: July 12, 2017 Outline AlphaGo Supervised Learning of Policy Networks

540 views • 15 slides
Reinforcement Learning CS 4100: Artificial Intelligence Reinforcement Learning Ja Jan-Wi

Reinforcement Learning CS 4100: Artificial Intelligence Reinforcement Learning Ja Jan-Wi

Reinforcement Learning CS 4100: Artificial Intelligence Reinforcement Learning Ja Jan-Wi Willem van de Meent Northeastern University [These slides were created by Dan Klein and Pieter Abbeel for CS188 Intro to AI at UC Berkeley. All CS188

145 views • 10 slides
Reinforcement Learning AIMA Chapters: 21.1, 21.2, 21.3. Sutton and Barto, Reinforcement Learning:

Reinforcement Learning AIMA Chapters: 21.1, 21.2, 21.3. Sutton and Barto, Reinforcement Learning:

Reinforcement Learning Reinforcement Learning AIMA Chapters: 21.1, 21.2, 21.3. Sutton and Barto, Reinforcement Learning: an Introduction, 2nd Edition: Chapters 6 (6.1 6.5) Outline Reinforcement Learning Reinforcement Learning: the

587 views • 27 slides
Module 11 Introduction to Reinforcement Learning CS 886 Sequential Decision Making and

Module 11 Introduction to Reinforcement Learning CS 886 Sequential Decision Making and

Module 11 Introduction to Reinforcement Learning CS 886 Sequential Decision Making and Reinforcement Learning University of Waterloo Machine Learning Supervised Learning Teacher tells learner what to remember Reinforcement Learning

401 views • 29 slides
1 Pure Reinforcement Learning vs. Reinforcement Learning Monte-Carlo Planning No knowledge

1 Pure Reinforcement Learning vs. Reinforcement Learning Monte-Carlo Planning No knowledge

CSE 573: Artificial Intelligence Todo Reinforcement Learning Add simulations from 473 Dan Weld Add UCB bound (cut bolzman &amp; constant epsilon Add snazzy videos (pendulum, zico kolter See http://www See http://www-

462 views • 11 slides
Safe Reinforcement Learning Philip S. Thomas Stanford CS234: Reinforcement Learning, Guest

Safe Reinforcement Learning Philip S. Thomas Stanford CS234: Reinforcement Learning, Guest

Safe Reinforcement Learning Philip S. Thomas Stanford CS234: Reinforcement Learning, Guest Lecture May 24, 2017 Lecture overview What makes a reinforcement learning algorithm safe ? Notation Creating a safe reinforcement learning

1.42k views • 88 slides
Larry Holder School of EECS Washington State University Artificial Intelligence 1 } Classic AI

Larry Holder School of EECS Washington State University Artificial Intelligence 1 } Classic AI

Larry Holder School of EECS Washington State University Artificial Intelligence 1 } Classic AI challenge Easy to represent Difficult to solve } Perfect information (e.g., Chess, Checkers) Fully observable and deterministic }

405 views • 27 slides
Intelligent Agents Chapter 2 Intelligent Agents p.1/25 Outline Agents and environments

Intelligent Agents Chapter 2 Intelligent Agents p.1/25 Outline Agents and environments

Intelligent Agents Chapter 2 Intelligent Agents p.1/25 Outline Agents and environments Rationality PEAS (Performance measure, Environment, Actuators, Sensors) Environment types Agent types Intelligent Agents p.2/25 Agents and

364 views • 23 slides
OPTICAL QUANTUM DOTS FOR QUANTUM INFORMATION Tom Reinecke Naval Research Laboratory Washington,

OPTICAL QUANTUM DOTS FOR QUANTUM INFORMATION Tom Reinecke Naval Research Laboratory Washington,

OPTICAL QUANTUM DOTS FOR QUANTUM INFORMATION Tom Reinecke Naval Research Laboratory Washington, DC , USA reinecke@nrl.navy.mil outline outline single spin qubits single spin qubits two qubit gates two qubit gates

696 views • 25 slides
Evalua&amp;onoftheSimulated PlanetaryBoundaryLayerin

Evalua&amp;onoftheSimulated PlanetaryBoundaryLayerin

Evalua&amp;onoftheSimulated PlanetaryBoundaryLayerin EasternTexas JennaKolling JonathanPleim(USEPA),WilliamVizuete(UNC),HarveyJeffries(UNC) October12,2010

228 views • 20 slides
A Desktop Can Machines Learn? Pascal Poupart Associate Professor David R. Cheriton School of

A Desktop Can Machines Learn? Pascal Poupart Associate Professor David R. Cheriton School of

A Desktop Can Machines Learn? Pascal Poupart Associate Professor David R. Cheriton School of Computer Science University of Waterloo 1 2 A Computer Program Machine Learning Arthur Samuel (1959): Machine learning is the field of study

367 views • 4 slides
CS440/ECE448 Lecture 12: Stochastic Games, Stochastic Search, and Learned Evaluation Functions

CS440/ECE448 Lecture 12: Stochastic Games, Stochastic Search, and Learned Evaluation Functions

CS440/ECE448 Lecture 12: Stochastic Games, Stochastic Search, and Learned Evaluation Functions Slides by Svetlana Lazebnik, 9/2016 Modified by Mark Hasegawa-Johnson, 2/2019 Types of game environments Deterministic Stochastic Perfect

600 views • 38 slides
POMDPs and Policy Gradients MLSS 2006, Canberra Douglas Aberdeen Canberra Node, RSISE Building

POMDPs and Policy Gradients MLSS 2006, Canberra Douglas Aberdeen Canberra Node, RSISE Building

POMDPs and Policy Gradients MLSS 2006, Canberra Douglas Aberdeen Canberra Node, RSISE Building Australian National University 15th February 2006 Outline Introduction 1 What is Reinforcement Learning? Types of RL Value-Methods 2 Model

849 views • 43 slides
Reinforcement Learning II George Konidaris gdk@cs.brown.edu Fall 2019 Reinforcement Learning

Reinforcement Learning II George Konidaris gdk@cs.brown.edu Fall 2019 Reinforcement Learning

Reinforcement Learning II George Konidaris gdk@cs.brown.edu Fall 2019 Reinforcement Learning t r t R = max : S A t =0 MDPs Agent interacts with an environment At each time t: Receives sensor signal s t

717 views • 40 slides

More recommend