Practical Bayesian Optimization of Machine Learning Algorithms

DSpace/Manakin Repository

Practical Bayesian Optimization of Machine Learning Algorithms

Citable link to this page

 

 
Title: Practical Bayesian Optimization of Machine Learning Algorithms
Author: Adams, Ryan Prescott; Snoek, Jasper; Larochelle, Hugo

Note: Order does not necessarily reflect citation order of authors.

Citation: Snoek, Jasper, Hugo Larochelle, and Ryan Prescott Adams. 2012. Practical Bayesian optimization of machine learning algorithms. Advances in Neural Information Processing Systems 25: 2960-2968
Full Text & Related Files:
Abstract: Machine learning algorithms frequently require careful tuning of model hyperparameters, regularization terms, and optimization parameters. Unfortunately, this tuning is often a "black art" that requires expert experience, unwritten rules of thumb, or sometimes brute-force search. Much more appealing is the idea of developing automatic approaches which can optimize the performance of a given learning algorithm to the task at hand. In this work, we consider the automatic tuning problem within the framework of Bayesian optimization, in which a learning algorithm's generalization performance is modeled as a sample from a Gaussian process (GP). The tractable posterior distribution induced by the GP leads to efficient use of the information gathered by previous experiments, enabling optimal choices about what parameters to try next. Here we show how the effects of the Gaussian process prior and the associated inference procedure can have a large impact on the success or failure of Bayesian optimization. We show that thoughtful choices can lead to results that exceed expert-level performance in tuning machine learning algorithms. We also describe new algorithms that take into account the variable cost (duration) of learning experiments and that can leverage the presence of multiple cores for parallel experimentation. We show that these proposed algorithms improve on previous automatic procedures and can reach or surpass human expert-level optimization on a diverse set of contemporary algorithms including latent Dirichlet allocation, structured SVMs and convolutional neural networks.
Other Sources: http://arxiv.org/abs/1206.2944
Terms of Use: This article is made available under the terms and conditions applicable to Open Access Policy Articles, as set forth at http://nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#OAP
Citable link to this page: http://nrs.harvard.edu/urn-3:HUL.InstRepos:11708816
Downloads of this work:

Show full Dublin Core record

This item appears in the following Collection(s)

 
 

Search DASH


Advanced Search
 
 

Submitters