Analysis reveals necessity of adaptive regularisers in online convex optimization, suggesting fixed regularisation fails in certain dimension regimes.
We study online convex optimization on ₚ-balls in Rᵈ for $p > 2$. While always sub-linear, the optimal regret exhibits a shift between the high-dimensional setting ($d > T$), when the dimension d is greater than the time horizon T and the low-dimensional setting (d ≤ T). We show that Follow-the-Regularised-Leader (FTRL) with time-varying regularisation which is adaptive to the dimension regime is anytime optimal for all dimension regimes. Motivated by this, we ask whether it is possible to obtain anytime optimality of FTRL with fixed non-adaptive regularisation. Our main result establishes that for separable regularisers, adaptivity in the regulariser is necessary, and that any fixed regulariser will be sub-optimal in one of the two dimension regimes. Finally, we provide lower bounds which rule out sub-linear regret bounds for the linear bandit problem in sufficiently high-dimension for all ₚ-balls with p ≥ 1.
No takes yet. Share an insight, caveat, or question.
Johnson et al. (2025) studied this question.