Synonyms
Definition
Error correcting output codes are an ensemble learning technique. It is applied to a problem with multiple classes, decomposing it into several binary problems. Each class is first encoded as a binary string of length T, assuming we have T models in the ensemble. Each model then tries to separate a subset of the original classes from all the others. For example, one model might learn to distinguish “class A” from “not class A.” After the predictions, with T models we have a binary string of length T. The class encoding that is closest to this binary string (using Hamming distance) is the final decision of the ensemble.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Recommended Reading
Kong EB, Dietterich TG (1995) Error-correcting output coding corrects bias and variance. In: International conference on machine learning, Tahoe City
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer Science+Business Media New York
About this entry
Cite this entry
(2017). Error Correcting Output Codes. In: Sammut, C., Webb, G.I. (eds) Encyclopedia of Machine Learning and Data Mining. Springer, Boston, MA. https://doi.org/10.1007/978-1-4899-7687-1_260
Download citation
DOI: https://doi.org/10.1007/978-1-4899-7687-1_260
Published:
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4899-7685-7
Online ISBN: 978-1-4899-7687-1
eBook Packages: Computer ScienceReference Module Computer Science and Engineering