Error-correcting output codes with ensemble diversity for robust learning in neural networks

• Main Authors: Song, Yang, Kang, Qiyu, Tay, Wee Peng
• Other Authors: School of Electrical and Electronic Engineering
• Format: Conference or Workshop Item
• Language: English
• Published: 2021
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence; Deep Learning; Neural Networks
• Online Access: https://hdl.handle.net/10356/147336
• Institution: Nanyang Technological University

Though deep learning has been applied successfully in many scenarios, malicious inputs with human-imperceptible perturbations can make it vulnerable in real applications. This paper proposes an error-correcting neural network (ECNN) that combines a set of binary classifiers to combat adversarial exam...