Evaluating the Performance of Explanation Methods on Ordinal Regression CNN Models

• Type: Article in International Conference Proceedings Book
• Date: 2023

Authors

Barbero-Gómez, J (Author) Other; Cruz, R (Author) FEUP; Jaime S Cardoso (Author) FEUP; Gutiérrez, PA (Author) Other; Hervás-Martínez, C (Author) Other

Conference proceedings International

Title: ADVANCES IN COMPUTATIONAL INTELLIGENCE, IWANN 2023, PT II

Pages: 529-540

17th International Work-Conference on Artificial Neural Networks (IWANN)

Ponta Delgada, PORTUGAL, JUN 19-21, 2023

Indexing

ISI Web of Knowledge - 0 Citations

Scopus - 0 Citations

Other information

Authenticus ID: P-00Z-3F2

DOI: 10.1007/978-3-031-43078-7_43

Abstract (EN): This paper introduces an evaluation procedure to validate the efficacy of explanation methods for Convolutional Neural Network (CNN) models in ordinal regression tasks. Two ordinal methods are contrasted against a baseline using cross-entropy, across four datasets. A statistical analysis demonstrates that attribution methods, such as Grad-CAM and IBA, perform significantly better when used with ordinal regression CNN models compared to a baseline approach in most ordinal and nominal metrics. The study suggests that incorporating ordinal information into the attribution map construction process may improve the explanations further.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 12