Laboratory of Computer and Information Science / Neural Networks Research Centre CIS Lab Helsinki University of Technology

Icasso: software for investigating the reliability of ICA estimates by clustering and visualization

A major problem in application of independent component analysis (ICA) is that the reliability of the estimated independent components is not known. Firstly, the finite sample size induces statistical errors in the estimation. Secondly, as real data never exactly follows the ICA model, the contrast function used in the estimation may have many local minima which are all equally good, or the practical algorithm may not always perform properly, for example getting stuck in local minima with strongly suboptimal values of the contrast function.

We present an explorative visualization method for investigating the relations between estimates from FastICA. The algorithmic and statistical reliability/stability is investigated by running the algorithm many times with different initial values or with differently bootstrapped data sets, respectively.

Reliable estimates correspond to tight clusters, and unreliable ones to points which do not belong to any such cluster. We have developed a software package called Icasso to implement these operations for assessment of ICA estimates.

Documentation and references:

Icasso 1.21 software package runs under MATLAB. At least version 6.1 is required. It needs also the FastICA software package.

Read about Icasso (copyright notice) and download the software.

If you have any comments or bug reports on the package, contact Aapo Hyvärinen.

You are at: CIS → Icasso

Page maintained by, last updated Tuesday, 05-Mar-2013 13:51:53 EET