Home › Categories › Technology› Http://ceur-ws.org/V Ol-2473/paper4.pdf

Http://ceur-ws.org/V Ol-2473/paper4.pdf

paper4.pdf

🔗 Source: ceur-ws.org

📊 Size: 429 KB

📄 Pages: 10 pages

⬇️ Downloads: 80

Summary

Rules extraction from neural networks trained on multimedia data is explored, focusing on methods that extract logical rules from trained networks. Three methods, ANN-DT, DeepRED, and HypInv, are applied to recognized text data and speech acoustic data with different subsets of features. The performance of these methods is compared, and statistical analysis is performed to evaluate the quality of the extracted rules, including accuracy, fidelity, and comprehensibility. The study aims to demonstrate the applicability of these methods to multi-class data with high input dimension, providing a proof of concept for their use in multimedia data analysis.

Download File

Description

Rules extraction from neural networks trained on multimedia data is explored, focusing on methods that extract logical rules from trained networks.

Technical Information

File Format: PDF
File Size: 429 KB
Pages: 10
Language: EN
Total Downloads: 80
Last Updated: 6 days ago

Document Overview

This PDF document about http://ceur-ws.org/V ol-2473/paper4.pdf provides comprehensive information and guidance. Whether you're a beginner or advanced user, this resource offers valuable insights into http://ceur-ws.org/V ol-2473/paper4.pdf.