Skip to content

Reinforcement learning exercises from Sutton & Barto's "Reinforcement Learning: An Introduction" (2014)

Notifications You must be signed in to change notification settings

josephlee188/RL_Exercises

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

RL_Exercises

Reinforcement learning exercises from R.S. Sutton & A. Barto's "Reinforcement Learning: An Introduction" (1992)

Jack's car rental problem

Finding optimal strategy for Jack which gives optimal reward (please refer to the book for details of the problem).

Day 0Day 1Day 2 Day 3 Day 4 Day 5

where the heatmaps are through Day 0 ~ 5.

$n_1$: Number of cars at parking lot 1.

$n_2$: Number of cars at parking lot 2.

The colors represent the number of cars to be moved from lot 1 to 2.

 

Windy gridworld

Uses Sarsa on-policy TD algorithm to find the quickest route to the goal when wind is blowing upwards.

windygw

The color represents steps.

 

Mountain car problem

Using TD($\lambda$) with continuous state (discrete action) to find optimal policy for car to reach the goal.

After 10 episodes

10

After 100 episodes

100

After 1000 episodes

1000

Value function (after 100 episodes)

vf_100

About

Reinforcement learning exercises from Sutton & Barto's "Reinforcement Learning: An Introduction" (2014)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Languages