https://github.com/kraftpunk97-zz/gamblers-problem
A dynamic programming solution to the classic gambler's problem introduced in Sutton and Barton's RL book
https://github.com/kraftpunk97-zz/gamblers-problem
dynamic-programming gambler-problem reinforcement-learning-algorithms
Last synced: 3 months ago
JSON representation
A dynamic programming solution to the classic gambler's problem introduced in Sutton and Barton's RL book
- Host: GitHub
- URL: https://github.com/kraftpunk97-zz/gamblers-problem
- Owner: kraftpunk97-zz
- Created: 2019-04-07T09:56:12.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2019-04-07T10:04:08.000Z (about 6 years ago)
- Last Synced: 2025-01-10T01:36:22.753Z (4 months ago)
- Topics: dynamic-programming, gambler-problem, reinforcement-learning-algorithms
- Language: Python
- Size: 33.2 KB
- Stars: 1
- Watchers: 0
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# gamblers-problem
A dynamic programming solution to the classic gambler's problem introduced in Sutton and Barton's RL book.
Problem statement - The gambler has a stake s between 0 and 100. At each play he wagers an integer <= s. He wins that much with prob p, else he loses that much. If he builds his stake to 100 he wins (thus he never wagers more than (- 100 s)); if his stake falls to 0 he loses.
The solution (Or rather one of the possible solutions. It all boils down to the finer details, like how ties are broken the compiler etc...)
