Change search
ReferencesLink to record
Permanent link

Direct link
Regression conformal prediction with random forests
Stockholm University, Faculty of Social Sciences, Department of Computer and Systems Sciences.
2014 (English)In: Machine Learning, ISSN 0885-6125, E-ISSN 1573-0565, Vol. 97, no 1-2, 155-176 p.Article in journal (Refereed) Published
Abstract [en]

Regression conformal prediction produces prediction intervals that are valid, i.e., the probability of excluding the correct target value is bounded by a predefined confidence level. The most important criterion when comparing conformal regressors is efficiency; the prediction intervals should be as tight (informative) as possible. In this study, the use of random forests as the underlying model for regression conformal prediction is investigated and compared to existing state-of-the-art techniques, which are based on neural networks and k-nearest neighbors. In addition to their robust predictive performance, random forests allow for determining the size of the prediction intervals by using out-of-bag estimates instead of requiring a separate calibration set. An extensive empirical investigation, using 33 publicly available data sets, was undertaken to compare the use of random forests to existing state-of-the-art conformal predictors. The results show that the suggested approach, on almost all confidence levels and using both standard and normalized nonconformity functions, produced significantly more efficient conformal predictors than the existing alternatives.

Place, publisher, year, edition, pages
2014. Vol. 97, no 1-2, 155-176 p.
Keyword [en]
Conformal prediction, Random forests, Regression
National Category
Information Systems, Social aspects
URN: urn:nbn:se:su:diva-107969DOI: 10.1007/s10994-014-5453-0ISI: 000341431300008OAI: diva2:754242


Available from: 2014-10-09 Created: 2014-10-06 Last updated: 2014-10-09Bibliographically approved

Open Access in DiVA

No full text

Other links

Publisher's full text

Search in DiVA

By author/editor
Boström, Henrik
By organisation
Department of Computer and Systems Sciences
In the same journal
Machine Learning
Information Systems, Social aspects

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Altmetric score

Total: 17 hits
ReferencesLink to record
Permanent link

Direct link