Frugal Optimization for Cost-related Hyperparameters
This addresses the need for low-cost HPO to democratize machine learning, offering a novel method for a known bottleneck in controlling costs during optimization.
The paper tackles the problem of hyperparameter optimization (HPO) ignoring training costs, developing a cost-frugal solution with a randomized direct-search method that achieves a convergence rate of O(√d/√K) and an O(dε⁻²) cost approximation guarantee, and shows strong empirical results on AutoML benchmarks.
The increasing demand for democratizing machine learning algorithms calls for hyperparameter optimization (HPO) solutions at low cost. Many machine learning algorithms have hyperparameters which can cause a large variation in the training cost. But this effect is largely ignored in existing HPO methods, which are incapable to properly control cost during the optimization process. To address this problem, we develop a new cost-frugal HPO solution. The core of our solution is a simple but new randomized direct-search method, for which we prove a convergence rate of $O(\frac{\sqrt{d}}{\sqrt{K}})$ and an $O(dε^{-2})$-approximation guarantee on the total cost. We provide strong empirical results in comparison with state-of-the-art HPO methods on large AutoML benchmarks.