Error-correcting output codes in the framework of deep ordinal classification

Hits: 8191
Áreas de investigación:
Año:
2021
Tipo de publicación:
Artículo en conferencia
Palabras clave:
Artificial Neural Networks, Ordinal Classification
Autores:
Volumen:
12862
Título del libro:
2021 International Work-conference on Artificial Neural Networks (IWANN 2021)
Número:
Part II
Serie:
Lecture Notes in Computer Science (LNCS)
Páginas:
3-13
Organización:
Online
Mes:
16nd-18th June
ISBN:
978-3-030-85029-6
ISSN:
0302-9743
BibTex:
Abstract:
Automatic classification tasks have been revolutionized by Convolutional Neural Networks (CNNs), but the focus has been on binary and nominal classification tasks. Only recently, ordinal classification (where class labels present a natural ordering) has been tackled through the framework of CNNs, such as adapting the classic Proportional Odds Model to deep architectures. Also, ordinal classification datasets commonly present a high imbalance in the number of samples of each class, making it an even harder problem. In this work, we present a new CNN architecture based on the Ordinal Binary Decomposition (OBD) technique using Error-Correcting Output Codes (ECOC) and show how it can improve performance over previously proposed methods.
Back