adzcai/incsfa

Incremental Slow Feature Analysis

================================== Incremental Slow Feature Analysis

Author - Varun Raj Kompella, IDSIA, Switzerland. (www.idsia.ch/~kompella)

This is a free software; you can redistribute it and/or modify it. The code is distributed in the hope that it will be useful.

If you plan to use this code in your research please cite: V. R. Kompella, M. Luciw and J. Schmidhuber. "Incremental Slow Feature Analysis: Adaptive Low-Complexity Slow Feature Updating from High-Dimensional Input Streams", Neural Computation Journal, Vol. 24 (11), pp. 2994--3024, 2012.

Abstract

Extract slowly varying components from the input data incrementally. More information about Incremental Slow Feature Analysis can be found in:

Kompella V.R, Luciw M. and Schmidhuber J., Incremental Slow Feature Analysis: Adaptive Low-Complexity Slow Feature Updating from High-Dimensional Input Streams, Neural Computation, 2012. (http://www.idsia.ch/~kompella/mywork/incsfa.html)

Some of the terminology used in the code is inspired from MDP toolkit (www.mdp-toolkit.sourceforge.net)

Files

• ccipca.py : Candid Covariance-Free Incremental PCA module
• mca.py : Minor Component Analysis module
• incsfa.py : Incremental Slow Feature Analysis module
• signalstats.py : Incremental signal stats modules
• trainer.py : trainer module used for training the modules (modes: 'Incremental', 'BlockIncremental', 'Batch')
• test_incsfa.py : Test example code for IncSFA

Optional Dependencies

• MDP toolkit for test_incsfa
• PyQTGraph for fast animated plotting (default is matplotlib)