We exploit a recently derived inversion scheme for arbitrary deep neural networks to develop a new semi-supervised learning framework that applies to a wide range of systems and problems. The approach outperforms current state-of-the-art methods on MNIST reaching 99.14% of test set accuracy while using 5 labeled examples per class. Experiments with one-dimensional signals highlight the generality of the method. Importantly, our approach is simple, efficient, and requires no change in the deep network architecture.
@article{arxiv.1711.04313,
title = {Semi-Supervised Learning via New Deep Network Inversion},
author = {Randall Balestriero and Vincent Roger and Herve G. Glotin and Richard G. Baraniuk},
journal= {arXiv preprint arXiv:1711.04313},
year = {2017}
}
Comments
arXiv admin note: substantial text overlap with arXiv:1710.09302