Feature Purification: How Adversarial Training Performs Robust Deep Learning.

Zeyuan Allen-Zhu Yuanzhi Li

Published in: FOCS (2021)

Keyphrases

deep learning
deep architectures
restricted boltzmann machine
unsupervised feature learning
unsupervised learning
machine learning
data sets
training set
feature vectors
supervised learning
mental models
deep belief networks
training samples
weakly supervised
viewpoint
image features