AI & ML
Cancellation-Free Constrained SquareCB for Safe Controller Selection
arXiv stat.ML0 views
arXiv:2611.21880v1 Announce Type: cross Abstract: We study constrained contextual bandits, in which a learner chooses one of $K$ actions in each round to minimize loss while keeping the expected cost of its play below a budget. Existing reductions to online regression weight the cost by a Lagrange multiplier learned from past constraint residuals, so slack in some rounds can offset violations in others. We instead run SquareCB proposed in \citep{foster2020beyond} with a rectified weight on the cost: it is zero when the estimated cost is within the budget, and otherwise just large enough to bring it back to the budget. Under realizability and Slater's condition with margin $\delta$, the algorithm attains regret $O(\sqrt{KT\,\mathrm{Reg}_{\mathsf{Sq}}(T)}/\delta)$ and cancellation-free violation $\sum_t(V_c^{p_t}-b)^+=O(\sqrt{KT\,\mathrm{Reg}_{\mathsf{Sq}}(T)})$, where $\mathrm{Reg}_{\mathsf{Sq}}$ is the error of the regression oracle. The same bounds hold when the budget must be met in every context, potentially adversarially chosen contexts. In controller selection for a safe-control-gym quadrotor and a BOPTEST building, the violation is 6 to 9 times lower than that of primal-dual and virtual-queue baselines, with a lower or comparable regret.
Source:https://arxiv.org/abs/2611.21880PreviousAdaptive Memory for Online Reinforcement Learning under Sparse Model ChangesNextDEFTS: Data-Efficient, Training-Free Few-Shot Learning for CLIP in the Text Subspace
Related posts

MacPaw's Moonlock Cybersecurity App Launches for iPhone
0MacRumors
Grand theft auto VI is an astonishing by American epic
0Hacker News New
Show HN: Globebug – A calm, conflict-safe travel OS for multi-city itineraries
0Hacker News New
Ask HN: Are you sick of buying domains
0Hacker News New
One missing, 11 injured after ADNOC vessel attacked in Strait of Hormuz
0Hacker News New
Show HN: Sidekick AI – A local-first VS Code extension with hybrid vector search
0Hacker News Show