A preliminary study of ordinal metrics to guide a multi-objective evolutionary algorithm

There are many metrics available to measure the goodness of a classifier when working with ordinal datasets. These measures are divided into product-moment and association metrics. In this paper, the behavior of several metrics is studied in different situations. In addition, two new measures associ...

Full description

Saved in:
Bibliographic Details
Published in2011 11th International Conference on Intelligent Systems Design and Applications pp. 1176 - 1181
Main Authors Cruz-Ramirez, M., Hervas-Martinez, C., Sanchez-Monedero, J., Gutierrez, P. A.
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.11.2011
Subjects
Analytical models
Correlation
Evolutionary computation
Intelligent systems
mean absolute error
Measurement uncertainty
multi-objective evolutionary algorithm
ordinal measures
Training
Online AccessGet full text
ISSN2164-7143
DOI10.1109/ISDA.2011.6121818

Cover

More Information
Summary:There are many metrics available to measure the goodness of a classifier when working with ordinal datasets. These measures are divided into product-moment and association metrics. In this paper, the behavior of several metrics is studied in different situations. In addition, two new measures associated with an ordinal classifier are defined: the maximum and the minimum mean absolute error of all the classes. From the results of this comparison, a pair of metrics is selected (one associated to the overall error and another one to the error of the class with lowest level of classification) to guide the evolution of a multi-objective evolutionary algorithm, obtaining good results in generalization on ordinal datasets.
ISSN:2164-7143
DOI:10.1109/ISDA.2011.6121818