Cogprints

Behaviourally meaningful representations from normalisation and context-guided denoising

Valpola, Harri (2004) Behaviourally meaningful representations from normalisation and context-guided denoising. [Departmental Technical Report]

Full text available as:

[img]
Preview
PDF
142Kb

Abstract

Many existing independent component analysis algorithms include a preprocessing stage where the inputs are sphered. This amounts to normalising the data such that all correlations between the variables are removed. In this work, I show that sphering allows very weak contextual modulation to steer the development of meaningful features. Context-biased competition has been proposed as a model of covert attention and I propose that sphering-like normalisation also allows weaker top-down bias to guide attention.

Item Type:Departmental Technical Report
Subjects:Neuroscience > Computational Neuroscience
Computer Science > Machine Learning
Computer Science > Neural Nets
Computer Science > Artificial Intelligence
ID Code:3633
Deposited By:Valpola, Harri
Deposited On:14 May 2004
Last Modified:11 Mar 2011 08:55

Metadata

Repository Staff Only: item control page