Learn · test · build

Multi-armed bandits,
from intuition to Python.

Explore decision-making algorithms through interactive challenges, then use the same ideas in reliable, reproducible Python experiments.

Interactive learning

Enter PyMAB Arcade

Explore all 27 concrete PyMAB policies across seven game worlds. Adjust policy settings and reward environments, then inspect each Python decision.

Reference and guides

Install PyMAB, design simulations, compare policies, evaluate logged data, and explore the complete Python API.