Next:
``Nongaussian is independent''
Up:
Independent Component Analysis: A
Previous:
Why Gaussian variables are
Principles of ICA estimation
``Nongaussian is independent''
Measures of nongaussianity
Kurtosis
Negentropy
Approximations of negentropy
Minimization of Mutual Information
Mutual Information
Defining ICA by Mutual Information
Maximum Likelihood Estimation
The likelihood
The Infomax Principle
Connection to mutual information
ICA and Projection Pursuit
Aapo Hyvarinen
2000-04-19
