Venue: NeurIPS
Year: 2020
Paper: https://proceedings.neurips.cc/paper/2020/hash/c7af0926b294e47e52e46cfebe173f20-Abstract.html
Abstract
Selecting optimal hyperparameters is a key challenge in machine learning. An exciting recent result showed it is possible to learn high-performing hyperparameter schedules on the fly in a single training run through methods inspired by Evolutionary Algorithms. These approaches have been shown to increase performance across a wide variety of machine learning tasks, ranging from supervised (SL) to reinforcement learning (RL). However, since they remain primarily evolutionary, they act in a greedy fashion, thus require a combination of vast computational resources and carefully selected meta-parameters to effectively explore the hyperparameter space. To address these shortcomings we look to Bayesian Optimization (BO), where a Gaussian Process surrogate model is combined with an acquisition function to produce a principled mechanism to trade off exploration vs exploitation. Our approach, which we call Probabilistic Population-Based Training (P2BT), is able to transfer sample efficiency of BO to the online setting, making it possible to achieve these traits in a single training run. We show that P2BT is able to achieve high performance with only a small population size, making it useful for all researchers regardless of their computational resources.
Additional information
PBT: https://deepmind.com/blog/article/population-based-training-neural-networks
