Numerical variable reconstruction from ordinal categories based on probability distributions

Research areas:

• Methodology - Machine learning

Year:

2011

Type of Publication:

In Proceedings

Keywords:

ordinal classification, ordinal regression, support vector machine, neural networks

Authors:

• Sánchez-Monedero, Javier
• Carbonero-Ruz, Mariano
• Becerra-Alonso, David
• Martínez-Estudillo, Francisco José
• Gutiérrez, Pedro Antonio
• Hervás-Martínez, César

Book title:

11th International Conference on Intelligent Systems Design andApplications (ISDA 2011)

Pages:

1182-1187

Address:

Cordoba, Spain, Spain

Month:

November

BibTex:

@conference{Sanchez2011isda, author = "Javier S{\'a}nchez-Monedero and Mariano Carbonero-Ruz and David Becerra-Alonso and Francisco Jos{\'e} Mart{\'i}nez-Estudillo and Pedro Antonio Guti{\'e}rrez and C{\'e}sar Herv{\'a}s-Mart{\'i}nez", abstract = "Ordinal classification problems are an active research area in the machine learning community. Many previous works adapted state-of-art nominal classifiers to improve ordinal classification so that the method can take advantage of the ordinal structure of the dataset. However, these method improvements often rely upon a complex mathematical basis and they usually belong to the training algorithm and model. This paper presents a novel method for generally adapting classification and regression models, such as artificial neural networks or support vector machines. The ordinal classification problem is reformulated as a regression problem by the reconstruction of a numeric variable which represents the different ordered class labels. Despite the simplicity and generality of the method, results are competitive in comparison with very specific methods for ordinal regression.", address = "Cordoba, Spain, Spain", booktitle = "11th International Conference on Intelligent Systems Design andApplications (ISDA 2011)", keywords = "ordinal classification, ordinal regression, support vector machine, neural networks", month = "nov", pages = "1182-1187", title = "{N}umerical variable reconstruction from ordinal categories based on probability distributions", year = "2011", }

Abstract:

Ordinal classification problems are an active research area in the machine learning community. Many previous works adapted state-of-art nominal classifiers to improve ordinal classification so that the method can take advantage of the ordinal structure of the dataset. However, these method improvements often rely upon a complex mathematical basis and they usually belong to the training algorithm and model. This paper presents a novel method for generally adapting classification and regression models, such as artificial neural networks or support vector machines. The ordinal classification problem is reformulated as a regression problem by the reconstruction of a numeric variable which represents the different ordered class labels. Despite the simplicity and generality of the method, results are competitive in comparison with very specific methods for ordinal regression.