A Robust Minimal Learning Machine based on the M-Estimator
Abstract
In this paper we propose a robust Minimal Learning Machine
(R-RLM) for regression problems. The proposed method uses a robust
M-estimator to generate a linear mapping between input and output
distances matrices of MLM. The R-MLM was tested on one synthetic
and three real world datasets that were contaminated with an increasing
number of outliers. The method achieved a performance comparable to
the robust Extreme Learning Machine (R-RLM) and thus can be seen as
a valid alternative for regression tasks on datasets with outliers.
Main Authors
Format
Conferences
Conference paper
Published
2017
Subjects
Publication in research information system
Publisher
ESANN
Original source
https://www.elen.ucl.ac.be/Proceedings/esann/esannpdf/es2017-44.pdf
The permanent address of the publication
https://urn.fi/URN:NBN:fi:jyu-201805162639Use this for linking
Parent publication ISBN
978-2-87587-039-1
Review status
Peer reviewed
Conference
European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning
Language
English
Is part of publication
ESANN 2017 : Proceedings of the 25th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning
Citation
- Kärkkäinen, T., Gomes, J., Mesquita, D., Freire, A., & Junior, A. S. (2017). A Robust Minimal Learning Machine based on the M-Estimator. In ESANN 2017 : Proceedings of the 25th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (pp. 383-388). ESANN. https://www.elen.ucl.ac.be/Proceedings/esann/esannpdf/es2017-44.pdf
License
Copyright© the Authors, 2017.