Cumulative link models for deep ordinal classification

Research areas:

• Methodology - Ordinal classification

Year:

2020

Type of Publication:

Article

Keywords:

Deep learning, Ordinal regression, Cumulative link models, Kappa index

Authors:

• Vargas, Víctor Manuel
• Gutiérrez, Pedro Antonio
• Hervás-Martínez, César

Journal:

Neurocomputing

Volume:

401

Pages:

48-58

Month:

August

ISSN:

0925-2312

BibTex:

@article{Vargas2020Neucom, author = "V{\'i}ctor Manuel Vargas and Pedro Antonio Guti{\'e}rrez and C{\'e}sar Herv{\'a}s-Mart{\'i}nez", abstract = "This paper proposes a deep convolutional neural network model for ordinal regression by considering a family of probabilistic ordinal link functions in the output layer. The link functions are those used for cumulative link models, which are traditional statistical linear models based on projecting each pattern into a 1-dimensional space. A set of ordered thresholds splits this space into the different classes of the problem. In our case, the projections are estimated by a non-linear deep neural network. To further improve the results, we combine these ordinal models with a loss function that takes into account the distance between the categories, based on the weighted Kappa index. Three different link functions are studied in the experimental study, and the results are contrasted with statistical analysis. The experiments run over two different ordinal classification problems and the statistical tests confirm that these models improve the results of a nominal model and outperform other robust proposals considered in the literature.", awards = "JCR(2020): 5.719 Position: 30/140 (Q1) Category: COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE.", comments = "JCR(2020): 5.719 Position: 30/140 (Q1) Category: COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE.", doi = "10.1016/j.neucom.2020.03.034", issn = "0925-2312", journal = "Neurocomputing", keywords = "Deep learning, Ordinal regression, Cumulative link models, Kappa index", month = "August", note = "JCR(2020): 5.719 Position: 30/140 (Q1) Category: COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE.", pages = "48-58", title = "{C}umulative link models for deep ordinal classification", url = "doi.org/10.1016/j.neucom.2020.03.034", volume = "401", year = "2020", }

Note:

JCR(2020): 5.719 Position: 30/140 (Q1) Category: COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE.

Abstract:

This paper proposes a deep convolutional neural network model for ordinal regression by considering a family of probabilistic ordinal link functions in the output layer. The link functions are those used for cumulative link models, which are traditional statistical linear models based on projecting each pattern into a 1-dimensional space. A set of ordered thresholds splits this space into the different classes of the problem. In our case, the projections are estimated by a non-linear deep neural network. To further improve the results, we combine these ordinal models with a loss function that takes into account the distance between the categories, based on the weighted Kappa index. Three different link functions are studied in the experimental study, and the results are contrasted with statistical analysis. The experiments run over two different ordinal classification problems and the statistical tests confirm that these models improve the results of a nominal model and outperform other robust proposals considered in the literature.

Comments:

JCR(2020): 5.719 Position: 30/140 (Q1) Category: COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE.