vdutell / FeedbackLCA Public

forked from dpaiton/FeedbackLCA

Notifications You must be signed in to change notification settings
Fork 0
Star 0

Semi-Supervised learning via the LCA sparse coding model with added Feedback connectivity

BSD-3-Clause license

0 stars 1 fork Branches Tags Activity

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 37 Commits
data		data
models		models
utils		utils
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
__init__.py		__init__.py
analyze_lca.py		analyze_lca.py
mnist_train.py		mnist_train.py

Repository files navigation

SemiSupervisedLCA

Modified LCA model to include feedback

TODO - New code

Add manually computed Entropy gradient to update_u equation
Plot incorrectly labeled examples from val or test set using save_data_tiled()
Histogram y_ values to determine relative scales
Write in normalization (regularization) method for classification 'w' matrix
Add alternative regularizers - leaky learning & dropout on 'phi' or 'w'

TODO - Code fixes

Fix plotting so that num_neurons does not have to have a sqrt
Fix auto-differentiation for LCAF inference

Copyright 2015 Yahoo Inc. Licensed under the terms of the New-BSD license. Please see LICENSE file in the project root for terms.

About

Semi-Supervised learning via the LCA sparse coding model with added Feedback connectivity

BSD-3-Clause license

Report repository

Releases

No releases published

Packages

No packages published

Languages

Python 100.0%