Daniel Blankenburg, Antonia Ellerbrock, Thomas Kesselheim et al.
We study the problem of minimizing an ordered norm of a load vector (indexed by a set of $d$ resources), where a finite number $n$ of customers $c$ contribute to the load of each resource by choosing a solution $x_c$ in a convex set $X_c \subseteq \mathbb{R}^d_{\geq 0}$; so we minimize $||\sum_{c}x_c||$ for some fixed ordered norm $||\cdot||$. We devise a randomized algorithm that computes a $(1+\varepsilon)$-approximate solution to this problem and makes, with high probability, $\mathcal{O}\left((n+d) (\varepsilon^{-2}+\log\log d)\log (n+d)\right)$ calls to oracles that minimize linear functions (with non-negative coefficients) over $X_c$. While this has been known for the $\ell_{\infty}$ norm via the multiplicative weights update method, existing proof techniques do not extend to arbitrary ordered norms. Our algorithm uses a resource price mechanism that is motivated by the follow-the-regularized-leader paradigm, and is expressed by smooth approximations of ordered norms. We need and show that these have non-trivial stability properties, which may be of independent interest. For each customer, we define dynamic cost budgets, which evolve throughout the algorithm, to determine the allowed step sizes. This leads to non-uniform updates and may even reject certain oracle solutions. Using non-uniform sampling together with a martingale argument, we can guarantee sufficient expected progress in each iteration, and thus bound the total number of oracle calls with high probability.