Understanding generalization and optimization performance of deep CNNs

• Main Authors: ZHOU, Pan, FENG, Jiashi
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2018
• Subjects: Theory and Algorithms
• Online Access: https://ink.library.smu.edu.sg/sis_research/9010; https://ink.library.smu.edu.sg/context/sis_research/article/10013/viewcontent/2018_ICML_deepCNNs.pdf
• Institution: Singapore Management University

This work aims to provide understandings on the remarkable success of deep convolutional neural networks (CNNs) by theoretically analyzing their generalization performance and establishing optimization guarantees for gradient descent based training algorithms. Specifically, for a CNN model consistin...