Coverage-Based Designs Improve Sample Mining and Hyperparameter Optimization

Gowtham Muniraju, Bhavya Kailkhura, Jayaraman J. Thiagarajan, Peer Timo Bremer, Cihan Tepedelenlioglu, Andreas Spanias

Research output: Contribution to journalArticlepeer-review

Abstract

Sampling one or more effective solutions from large search spaces is a recurring idea in machine learning (ML), and sequential optimization has become a popular solution. Typical examples include data summarization, sample mining for predictive modeling, and hyperparameter optimization. Existing solutions attempt to adaptively trade off between global exploration and local exploitation, in which the initial exploratory sample is critical to their success. While discrepancy-based samples have become the de facto approach for exploration, results from computer graphics suggest that coverage-based designs, e.g., Poisson disk sampling, can be a superior alternative. In order to successfully adopt coverage-based sample designs to ML applications, which were originally developed for 2-D image analysis, we propose fundamental advances by constructing a parameterized family of designs with provably improved coverage characteristics and developing algorithms for effective sample synthesis. Using experiments in sample mining and hyperparameter optimization for supervised learning, we show that our approach consistently outperforms the existing exploratory sampling methods in both blind exploration and sequential search with Bayesian optimization.

Original languageEnglish (US)
Article number9069306
Pages (from-to)1241-1253
Number of pages13
JournalIEEE Transactions on Neural Networks and Learning Systems
Volume32
Issue number3
DOIs
StatePublished - Mar 2021

Keywords

  • Coverage-based sample design
  • hyperparameter optimization
  • Poisson disk sampling (PDS)
  • predictive modeling
  • sequential optimization

ASJC Scopus subject areas

  • Software
  • Computer Science Applications
  • Computer Networks and Communications
  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'Coverage-Based Designs Improve Sample Mining and Hyperparameter Optimization'. Together they form a unique fingerprint.

Cite this