Semi-automatic CNN Architectural Pruning using the Bayesian Case Model and Dimensionality Reduction Visualization

Author(s)

Marcilio-Jr, Wilson E. ; Eler, Danilo ; Guilherme, Ivan ; Hurter, C ; Purchase, H ; Bouatouch, K

• Document type: Journal article
• Source: PROCEEDINGS OF THE 17TH INTERNATIONAL JOINT CONFERENCE ON COMPUTER VISION, IMAGING AND COMPUTER GRAPHICS THEORY AND APPLICATIONS (VISAPP), VOL 4; v. N/A, p. 7-pg., 2022-01-01.

Abstract

Visualization techniques have been applied to reasoning about complex machine learning models. These visual approaches aim to enhance the understanding of black-box models' decisions or guide in hyperparameters configuration, such as the number of layers and neurons/filters in deep neural networks. While several works address the architectural tuning of convolutional neural networks (CNNs), only a few works face the problem from a semi-automatic perspective. This work presents a novel application of the Bayesian Case Model that uses visualization strategies to convey the most important filters of convolutional layers for image classification. A heatmap coordinated with a scatterplot visualization emphasizes the filters with the most contribution to the CNN prediction. Our methodology is evaluated on a case study using the MNIST dataset. (AU)

• FAPESP's process: 18/25755-8 - Multilevel visual representation to assist the exploration of data sets
• Grantee: Wilson Estécio Marcílio Júnior
• Support Opportunities: Scholarships in Brazil - Technical Training Program - Technical Training
• FAPESP's process: 18/17881-3 - Multilevel visual representation to assist the exploration of data sets
• Grantee: Danilo Medeiros Eler
• Support Opportunities: Regular Research Grants