8000 GitHub - msohcw/hog-contest: Tabular Q-learner to play the game of Hog.
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

msohcw/hog-contest

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Hog Contest

This program develops a strategy to play the two-player game of Hog (CS61A Fall '16). The game is simulated in C++, and the strategy is learned over time using reinforcement learning, specifically, tabular Q-learning.

  • memory.in stores the latest strategy being trained against
  • memory.50000 is the strategy trained over 50000 epochs
  • viz.py creates a visualization of a given strategy
  • hog_rl.cpp compiles to hog, hog_cv and hog_in_to_py
    • hog is the main file, run this to train the strategy
    • hog_cv is for cross validating across strategies, to compare strategies directly
    • hog_in_to_py converts a .in file to a list.py file which is in the format of a python list. Append this to viz.py for visualization

About

Tabular Q-learner to play the game of Hog.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published
0