|
|
|
|
|
Monthly
288 pp. per issue
6 x 9, illustrated
Founded: 1989
ISSN 0899-7667
E-ISSN 1530-888X
2010 Impact Factor: 2.290
|
April 2008, Vol. 20, No. 4, Pages 1091-1117
Posted Online February 22, 2008.
(doi:10.1162/neco.2008.03-07-489)
© 2008 Massachusetts Institute of Technology
A Study on Neural Learning on Manifold Foliations: The Case of the Lie Group SU(3)Simone FioriDipartimento di Elettronica, Intelligenza Artificiale e Telecomunicazioni, Facoltà di Ingegneria, Università Politecnica delle Marche, I-60131 Ancona, Italy. fiori@deit.univpm.it
Learning on differential manifolds may involve the optimization of a function of many parameters. In this letter, we deal with Riemannian-gradient-based optimization on a Lie group, namely, the group of unitary unimodular matrices SU(3). In this special case, subalgebras of the associated Lie algebra su(3) may be individuated by computing pair-wise Gell-Mann matrices commutators. Subalgebras generate subgroups of a Lie group, as well as manifold foliation. We show that the Riemannian gradient may be projected over tangent structures to foliation, giving rise to foliation gradients. Exponentiations of foliation gradients may be computed in closed forms, which closely resemble Rodriguez forms for the special orthogonal group SO(3). We thus compare optimization by Riemannian gradient and foliation gradients. Cited byIgor Aizenberg. (2010) Periodic Activation Function and a Modified Learning Algorithm for the Multivalued Neuron. IEEE Transactions on Neural Networks 21:12, 1939-1949 Online publication date: 1-Dec-2010. CrossRef Simone Fiori, Toshihisa Tanaka. (2009) An Algorithm to Compute Averages on Matrix Lie Groups. IEEE Transactions on Signal Processing 57:12, 4734-4743 Online publication date: 1-Dec-2009. CrossRef S FIORI. (2008) Lie-group-type neural system learning by manifold retractions  . Neural NetworksOnline publication date: 2-Oct-2008. CrossRef
|
|
|
|
MIT Press Journals |
Subscribe |
Contact Us |
Search |
Privacy Statement |
Terms and Conditions
© 2012 The MIT Press
|
| Technology Partner - Atypon Systems, Inc. |
|