Various configurations of two layer networks.
To look at:
"What is the Best Multi-Stage Architecture for Object Recognition?"
http://yann.lecun.com/exdb/publis/pdf/jarrett-iccv-09.pdf
http://cs.nyu.edu/~koray/publis/code/randomc101.tar.gz
"An Analysis of Single-Layer Networks in Unsupervised Feature Learning"
http://www.cs.stanford.edu/~acoates/papers/CoatesLeeNg_nips2010_dlwkshp_singlelayer.pdf
code:
https://github.com/jhjin/kmeans-learning-torch
http://fastml.com/the-secret-of-the-big-guys/
https://github.com/zygmuntz/the-secret-of-the-big-guys
"Web-Scale K-Means Clustering"
http://www.eecs.tufts.edu/~dsculley/papers/fastkmeans.pdf
http://stackoverflow.com/questions/21447351/minibatchkmeans-parameters
"Hierarchical k-Means for Unsupervised Learning"
http://www.andrew.cmu.edu/user/hgifford/projects/k_means.pdf
"PCANet: A Simple Deep Learning Baseline for Image Classification?"
http://arxiv.org/pdf/1404.3606v2.pdf
Actually also two layer:
http://www.cs.utoronto.ca/~kriz/conv-cifar10-aug2010.pdf
"How far can you get with a modern face recognition test set using only simple features?"
http://coxlab.org/pdfs/pinto-dicarlo-cox-cvpr-2009-mkl.pdf
Title of the paper | URL | Acc | Code |
---|---|---|---|
An Analysis of Single-Layer Networks in Unsupervised Feature Learning | link | 00 | yes |
PCANet: A Simple Deep Learning Baseline for Image Classification? | link | 00 | yes |
What is the Best Multi-Stage Architecture for Object Recognition? | link | 00 | yes |