Library for multi-armed bandit selection strategies, including efficient deterministic implementations of Thompson sampling and epsilon-greedy. - View it on GitHub