Skip to content

Latest commit

 

History

29 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 

Repository files navigation

reinforce

A 'plug and play' reinforcement learning library in Python.

Infers a Markov Decision Process from data and solves for the optimal policy.

Implementation based on Andrew Ng's notes.

More information related to this project can be found here.

Example Usage

observations = [
  { 'state_transitions': [
      { 'state': 'low', 'action': 'climb', 'state_': 'mid' },
      { 'state': 'mid', 'action': 'climb', 'state_': 'high' },
      { 'state': 'high', 'action': 'sink', 'state_': 'mid' },
      { 'state': 'mid', 'action': 'sink', 'state_': 'low' },
      { 'state': 'low', 'action': 'sink', 'state_': 'bottom' }
    ],
    'reward': 0
  },
  { 'state_transitions': [
      { 'state': 'low', 'action': 'climb', 'state_': 'mid' },
      { 'state': 'mid', 'action': 'climb', 'state_': 'high' },
      { 'state': 'high', 'action': 'climb', 'state_': 'top' },
    ],
    'reward': 0
  }
]

trap_states = [
  {
    'state_transitions': [
      { 'state': 'bottom', 'action': 'sink', 'state_': 'bottom' },
      { 'state': 'bottom', 'action': 'climb', 'state_': 'bottom' }
    ],
    'reward': 0
  },
  {
    'state_transitions': [
      { 'state': 'top', 'action': 'sink', 'state_': 'top' },
      { 'state': 'top', 'action': 'climb', 'state_': 'top' },
    ],
    'reward': 1
  },
]

from learn import MarkovAgent
mark = MarkovAgent(observations + trap_states)
mark.learn()

print(mark.policy)
# {'high': 'climb', 'top': 'sink', 'bottom': 'sink', 'low': 'climb', 'mid': 'climb'}
# NOTE: policy in top and bottom states is chosen randomly (doesn't affect state)

About

Simple reinforcement learning in Python.

Resources

Stars

202 stars

Watchers

16 watching

Forks

Releases

Packages

Used by

Contributors

Languages