James L. Weaver
Developer Advocate
 


jweaver@pivotal.io
JavaFXpert.com

@JavaFXpert

Katharine Beaumont
Developer / Mathematician

 


kbe@voxxed.com
voxxed.com

@KatharineCodes

Machine Learning

Workshop: Part Four

Reinforcement Learning

(Let's dive in now)

@KatharineCodes  @JavaFXpert

Using BURLAP for Reinforcement Learning

@KatharineCodes  @JavaFXpert

Learning to Navigate a Grid World with Q-Learning

@KatharineCodes  @JavaFXpert

Rules of this Grid World

  • Agent may move left, right, up, or down (actions)
  • Reward is 0 for each move
  • Reward is 5 for reaching top right corner (terminal state)
  • Agent can't move into a wall or off-grid
  • Agent doesn't have a model of the grid world.  It must discover as it interacts.

Challenge: Given that there is only one state that gives a reward, how can the agent work out what actions will get it to the reward?

(AKA the credit assignment problem)

Goal of an episode is to maximize total reward

@KatharineCodes  @JavaFXpert

Visualizing training episodes

From BasicBehavior example in  https://github.com/jmacglashan/burlap_examples

@KatharineCodes  @JavaFXpert

This Grid World's MDP (Markov Decision Process)

In this example, all actions are deterministic

@KatharineCodes  @JavaFXpert

Agent learns optimal policy from interactions with the environment (s, a, r, s')

@KatharineCodes  @JavaFXpert

Expected future discounted rewards, and polices

@KatharineCodes  @JavaFXpert

This example used discount factor 0.9

Low discount factors cause agent to prefer immediate rewards

@KatharineCodes  @JavaFXpert

How often should the agent try new paths vs. greedily taking known paths?

@KatharineCodes  @JavaFXpert

Q-Learning approach to reinforcement learning

Left Right Up Down
...
2, 7 2.65 4.05 0.00 3.20
2, 8 3.65 4.50 4.50 3.65
2, 9 4.05 5.00 5.00 4.05
2, 10 4.50 4.50 5.00 3.65
...

Q-Learning table of expected values (cumulative discounted rewards) as a result of taking an action from a state and following an optimal policy.  Here's an explanation of how calculations in a Q-Learning table are performed.

Actions

States

@KatharineCodes  @JavaFXpert

Tic-Tac-Toe with Reinforcement Learning

Learning to win from experience rather than by being trained

@KatharineCodes  @JavaFXpert

Inspired by the Tic-Tac-Toe Example section...

@KatharineCodes  @JavaFXpert

Tic-Tac-Toe Learning Agent and Environment

X

O

Our learning agent is the "X" player, receiving +5 for winning, -5 for losing, and -1 for each turn

The "O" player is part of the Environment.  State and reward updates that it gives the Agent consider the "O" play.

@KatharineCodes  @JavaFXpert

Tic-Tac-Toe state is the game board and status

States 0 1 2 3 4 5 6 7 8
O I X I O X X I O, O won N/A N/A N/A N/A N/A N/A N/A N/A N/A
I  I  I  I  I  I O I X, in prog 1.24 1.54 2.13 3.14 2.23 3.32 N/A 1.45 N/A
I  I O I  I X O I X, in prog 2.34 1.23 N/A 0.12 2.45 N/A N/A 2.64 N/A
I  I O O X X O I X, in prog +4.0 -6.0 N/A N/A N/A N/A N/A -6.0 N/A
X I O I  I X O I X, X won N/A N/A N/A N/A N/A N/A N/A N/A N/A
...

Q-Learning table of expected values (cumulative discounted rewards) as a result of taking an action from a state and following an optimal policy

Actions (Possible cells to play)

Unoccupied cell represented with an I in the States column

@KatharineCodes  @JavaFXpert

Tic-Tac-Toe with Reinforcement Learning

@KatharineCodes  @JavaFXpert

Summary of neural network links (1/2)

@KatharineCodes  @JavaFXpert

Summary of neural network links (2/2)

@KatharineCodes  @JavaFXpert

Summary of reinforcement learning links

BURLAP library: http://burlap.cs.brown.edu

BURLAP examples including BasicBehavior:
https://github.com/jmacglashan/burlap_examples

Markov Decision Process:
https://en.wikipedia.org/wiki/Markov_decision_process

Q-Learning table calculations: http://artint.info/html/ArtInt_265.html

Exploitation vs. exploration:
https://en.wikipedia.org/wiki/Multi-armed_bandit

Reinforcement Learning: An Introduction:
https://webdocs.cs.ualberta.ca/~sutton/book/bookdraft2016sep.pdf

Tic-tac-toe reinforcement learning app:
https://github.com/JavaFXpert/tic-tac-toe-rl

@KatharineCodes  @JavaFXpert

James L. Weaver
Developer Advocate
 


jweaver@pivotal.io
JavaFXpert.com

@JavaFXpert

Katharine Beaumont
Developer / Speaker

 


kbe@voxxed.com
voxxed.com

@KatharineCodes

Machine Learning

Workshop: Part Four

Machine Learning Exposed Workshop: Part Four

By javafxpert

Machine Learning Exposed Workshop: Part Four

Part four of Machine Learning Exposed workshop. Shedding light on machine learning, being gentle with the math.

  • 2,453