A Whole New Ball Game: A Primal Accelerated Method for Matrix Games and Minimizing the Maximum of Smooth Functions
This work addresses optimization challenges in machine learning and game theory, offering incremental algorithmic improvements for large-scale problems.
The paper tackles the problem of minimizing the maximum of smooth functions, particularly for matrix games, by developing an algorithm that computes an ε-approximate solution with optimal gradient evaluations up to polylogarithmic factors for large n, achieving runtime improvements over existing first-order and interior point methods in specific regimes.
We design algorithms for minimizing $\max_{i\in[n]} f_i(x)$ over a $d$-dimensional Euclidean or simplex domain. When each $f_i$ is $1$-Lipschitz and $1$-smooth, our method computes an $ε$-approximate solution using $\widetilde{O}(n ε^{-1/3} + ε^{-2})$ gradient and function evaluations, and $\widetilde{O}(n ε^{-4/3})$ additional runtime. For large $n$, our evaluation complexity is optimal up to polylogarithmic factors. In the special case where each $f_i$ is linear -- which corresponds to finding a near-optimal primal strategy in a matrix game -- our method finds an $ε$-approximate solution in runtime $\widetilde{O}(n (d/ε)^{2/3} + nd + dε^{-2})$. For $n>d$ and $ε=1/\sqrt{n}$ this improves over all existing first-order methods. When additionally $d = ω(n^{8/11})$ our runtime also improves over all known interior point methods. Our algorithm combines three novel primitives: (1) A dynamic data structure which enables efficient stochastic gradient estimation in small $\ell_2$ or $\ell_1$ balls. (2) A mirror descent algorithm tailored to our data structure implementing an oracle which minimizes the objective over these balls. (3) A simple ball oracle acceleration framework suitable for non-Euclidean geometry.