Automatic Exploration of Machine Learning Experiments on OpenML
pith:OC3HZGOJ Add to your LaTeX paper
What is a Pith Number?\usepackage{pith}
\pithnumber{OC3HZGOJ}
Prints a linked pith:OC3HZGOJ badge after your title and writes the identifier into PDF metadata. Compiles on arXiv with no extra files. Learn more
read the original abstract
Understanding the influence of hyperparameters on the performance of a machine learning algorithm is an important scientific topic in itself and can help to improve automatic hyperparameter tuning procedures. Unfortunately, experimental meta data for this purpose is still rare. This paper presents a large, free and open dataset addressing this problem, containing results on 38 OpenML data sets, six different machine learning algorithms and many different hyperparameter configurations. Results where generated by an automated random sampling strategy, termed the OpenML Random Bot. Each algorithm was cross-validated up to 20.000 times per dataset with different hyperparameters settings, resulting in a meta dataset of around 2.5 million experiments overall.
This paper has not been read by Pith yet.
Forward citations
Cited by 1 Pith paper
-
Machine learning models for estimating counterfactuals in a single-arm inflammatory bowel disease study
ML models trained on external IFX data can predict counterfactual outcomes for ADA patients, yielding treatment effect estimates aligned with propensity score matching and showing no significant difference between treatments.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.