Efficient hyperparameter optimization by way of PAC-Bayes bound minimization

by   John J. Cherian, et al.

Identifying optimal values for a high-dimensional set of hyperparameters is a problem that has received growing attention given its importance to large-scale machine learning applications such as neural architecture search. Recently developed optimization methods can be used to select thousands or even millions of hyperparameters. Such methods often yield overfit models, however, leading to poor performance on unseen data. We argue that this overfitting results from using the standard hyperparameter optimization objective function. Here we present an alternative objective that is equivalent to a Probably Approximately Correct-Bayes (PAC-Bayes) bound on the expected out-of-sample error. We then devise an efficient gradient-based algorithm to minimize this objective; the proposed method has asymptotic space and time complexity equal to or better than other gradient-based hyperparameter optimization methods. We show that this new method significantly reduces out-of-sample error when applied to hyperparameter optimization problems known to be prone to overfitting.



There are no comments yet.


page 1

page 2

page 3

page 4


Using Known Information to Accelerate HyperParameters Optimization Based on SMBO

Automl is the key technology for machine learning problem. Current state...

Optimizing Large-Scale Hyperparameters via Automated Learning Algorithm

Modern machine learning algorithms usually involve tuning multiple (from...

Self-tuning networks:

Hyperparameter optimization can be formulated as a bilevel optimization ...

Weighted Random Search for Hyperparameter Optimization

We introduce an improved version of Random Search (RS), used here for hy...

Optimizing Millions of Hyperparameters by Implicit Differentiation

We propose an algorithm for inexpensive gradient-based hyperparameter op...

DEHB: Evolutionary Hyberband for Scalable, Robust and Efficient Hyperparameter Optimization

Modern machine learning algorithms crucially rely on several design deci...

Stability and Generalization of Bilevel Programming in Hyperparameter Optimization

Recently, the (gradient-based) bilevel programming framework is widely u...

Code Repositories

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.