BriefGPT.xyz
Jun, 2015
基于图结构数据的深度卷积网络
Deep Convolutional Networks on Graph-Structured Data
HTML
PDF
Mikael Henaff, Joan Bruna, Yann LeCun
TL;DR
本文介绍了一种将谱网络与图估计过程相结合的方法，用于构建小学习复杂度和处理没有强规律的领域的深层神经网络，并在大规模分类问题中测试并达到更好的性能。
Abstract
deep learning
's recent successes have mostly relied on
convolutional networks
, which exploit fundamental statistical properties of images, sounds and video data: the local stationarity and
→