Welcome to Lilian Besson’s “AlgoBandits” project documentation!¶
A research framework for Single and Multi-Players Multi-Arms Bandits (MAB) Algorithms: UCB, KL-UCB, Thompson... and MusicalChair, ALOHA, MEGA, rhoRand etc.
See more on the GitHub page for this project: https://naereen.github.io/AlgoBandits/. The project is also hosted on Inria GForge, and the documentation can be seen online at http://banditslilian.gforge.inria.fr/.
Bandit algorithms, Lilian Besson’s “AlgoBandits” project¶
- Bandit algorithms, Lilian Besson’s AlgoBandits project
- How to run the code ?
- Policy aggregation algorithms
- Multi-players simulation environment
- Fairness vs. unfairness
- Short documentation of the API
- 💥 TODO
- Some illustrations for this project
- Jupyter Notebooks 📓 by Naereen @ GitHub
- List of notebooks for AlgoBandits documentation
- Fast C versions of the utilities in
- A note on execution times, speed and profiling
- A better approach?
- An even better approach?
- UML diagrams
Indices and tables¶
- Module Index
- Python Class Index,
- Python Function Index,
- Python Method Index,
- Python Static Method Index,
- Python Attribute Index,
- Search Page.
Should you use bandits?
In 2015, Chris Stucchio advised against the use of bandits, in the context of improving A/B testings, opposed to his 2013 blog post in favor of bandits, also for A/B testings. Both articles are worth reading, but in this research we are not studying A/B testing, and it has been already proved how efficient bandit algorithms can be for real-world and simulated cognitive radio networks. (See for instance this article by Wassim Jouini, Christophe Moy and Jacques Palicot, [“Multi-armed bandit based policies for cognitive radio’s decision making issues”, W Jouini, D Ernst, C Moy, J Palicot 2009]).