A Light-Weight Multi-Objective Asynchronous Hyper-Parameter Optimizer
This is an incremental improvement for researchers and practitioners needing efficient hyper-parameter tuning in machine learning.
The paper tackles hyper-parameter optimization by developing a light-weight system that minimizes a scalar cost from multiple objectives and supports user-driven trade-offs, achieving robustness to simulation issues.
We describe a light-weight yet performant system for hyper-parameter optimization that approximately minimizes an overall scalar cost function that is obtained by combining multiple performance objectives using a target-priority-limit scalarizer. It also supports a trade-off mode, where the goal is to find an appropriate trade-off among objectives by interacting with the user. We focus on the common scenario where there are on the order of tens of hyper-parameters, each with various attributes such as a range of continuous values, or a finite list of values, and whether it should be treated on a linear or logarithmic scale. The system supports multiple asynchronous simulations and is robust to simulation stragglers and failures.