Data complexity meta-features for regression problems

In meta-learning, classification problems can be described by a variety of features, including complexity measures. These measures allow capturing the complexity of the frontier that separates the classes. For regression problems, on the other hand, there is a lack of such type of measures. This pap...

Full description

Saved in:
Bibliographic Details
Published in:Machine learning Vol. 107; no. 1; pp. 209 - 246
Main Authors: Lorena, Ana C., Maciel, Aron I., de Miranda, Péricles B. C., Costa, Ivan G., Prudêncio, Ricardo B. C.
Format: Journal Article
Language:English
Published: New York Springer US 01.01.2018
Springer Nature B.V
Subjects:
Artificial Intelligence
Complexity
Computer Science
Control
Datasets
Learning
Machine learning
Mechatronics
Natural Language Processing (NLP)
Regression analysis
Robotics
Simulation and Modeling
Special Issue on Metalearning and Algorithm Selection
Complexity measures
Meta-features
Meta-learning
ISSN:0885-6125, 1573-0565
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In meta-learning, classification problems can be described by a variety of features, including complexity measures. These measures allow capturing the complexity of the frontier that separates the classes. For regression problems, on the other hand, there is a lack of such type of measures. This paper presents and analyses measures devoted to estimate the complexity of the function that should fitted to the data in regression problems. As case studies, they are employed as meta-features in three meta-learning setups: (i) the first one predicts the regression function type of some synthetic datasets; (ii) the second one is designed to tune the parameter values of support vector regressors; and (iii) the third one aims to predict the performance of various regressors for a given dataset. The results show the suitability of the new measures to describe the regression datasets and their utility in the meta-learning tasks considered. In cases (ii) and (iii) the achieved results are also similar or better than those obtained by the use of classical meta-features in meta-learning.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0885-6125
1573-0565
DOI:10.1007/s10994-017-5681-1