A Generative Model for Deep Convolutional Learning
Machine Learning
2015-04-17 v1 Machine Learning
Neural and Evolutionary Computing
Abstract
A generative model is developed for deep (multi-layered) convolutional dictionary learning. A novel probabilistic pooling operation is integrated into the deep model, yielding efficient bottom-up (pretraining) and top-down (refinement) probabilistic learning. Experimental results demonstrate powerful capabilities of the model to learn multi-layer features from images, and excellent classification results are obtained on the MNIST and Caltech 101 datasets.
Cite
@article{arxiv.1504.04054,
title = {A Generative Model for Deep Convolutional Learning},
author = {Yunchen Pu and Xin Yuan and Lawrence Carin},
journal= {arXiv preprint arXiv:1504.04054},
year = {2015}
}
Comments
3 pages, 1 figure, ICLR workshop