Efficient Change-Point Detection for Tackling Piecewise-Stationary Bandits
About
We introduce GLR-klUCB, a novel algorithm for the piecewise iid non-stationary bandit problem with bounded rewards. This algorithm combines an efficient bandit algorithm, kl-UCB, with an efficient, parameter-free, changepoint detector, the Bernoulli Generalized Likelihood Ratio Test, for which we provide new theoretical guarantees of independent interest. Unlike previous non-stationary bandit algorithms using a change-point detector, GLR-klUCB does not need to be calibrated based on prior knowledge on the arms' means. We prove that this algorithm can attain a $O(\sqrt{TA \Upsilon_T\log(T)})$ regret in $T$ rounds on some "easy" instances, where A is the number of arms and $\Upsilon_T$ the number of change-points, without prior knowledge of $\Upsilon_T$. In contrast with recently proposed algorithms that are agnostic to $\Upsilon_T$, we perform a numerical study showing that GLR-klUCB is also very efficient in practice, beyond easy instances.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Multi-Armed Bandit | Yahoo! Day 2 | Average Time (s)560 | 7 | |
| Multi-Armed Bandit | Yahoo! Day 3 | Average Time (s)613 | 7 | |
| Multi-Armed Bandit | Yahoo! Day 4 | Avg Time (s)683 | 7 | |
| Multi-Armed Bandit | Yahoo! Day 5 | Average Latency (s)2.42e+3 | 7 | |
| Multi-Armed Bandit | Yahoo! Day 6 | Average Computational Time (s)707 | 7 | |
| Multi-Armed Bandit | Yahoo! Day 7 | Average Computational Time (s)1.53e+3 | 7 | |
| Multi-Armed Bandit | Yahoo! Day 8 | Average Latency (s)957 | 7 | |
| Multi-Armed Bandit | Yahoo! dataset Day 9 | Average Computational Time (s)971 | 7 |