Abstract | ||
---|---|---|
Independent Subspace Analysis (ISA) is an extension of Independent Component Analysis (ICA) that aims to linearly transform a random vector such as to render groups of its components mutually independent. A recently proposed fixed-point algorithm is able to locally perform ISA if the sizes of the subspaces are known, however global convergence is a serious problem as the proposed cost function has additional local minima. We introduce an extension to this algorithm, based on the idea that the algorithm converges to a solution, in which subspaces that are members of the global minimum occur with a higher frequency. We show that this overcomes the algorithm's limitations. Moreover, this idea allows a blind approach, where no a priori knowledge of subspace sizes is required. |
Year | DOI | Venue |
---|---|---|
2009 | 10.1007/978-3-642-00599-2_33 | ICA |
Keywords | Field | DocType |
blind approach,additional local minimum,independent component analysis,independent subspaces,global convergence,algorithm converges,unknown dimensions,fixed-point algorithm,proposed cost function,higher frequency,hierarchical extraction,global minimum,independent subspace analysis | Subspace topology,A priori and a posteriori,Minimum description length,Algorithm,Linear subspace,Maxima and minima,Multivariate random variable,Independent component analysis,Independence (probability theory),Mathematics | Conference |
Volume | ISSN | Citations |
5441 | 0302-9743 | 3 |
PageRank | References | Authors |
0.39 | 9 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
P. Gruber | 1 | 51 | 5.80 |
Harold W. Gutch | 2 | 45 | 4.60 |
Fabian J. Theis | 3 | 931 | 85.37 |