Inverse Problems in Learning from Data

0349057 - ÚI 2011 RIV PT eng C - Conference Paper (international conference)
Kůrková, Věra
Inverse Problems in Learning from Data.
ICNC 2010. Proceedings of the International Conference on Neural Computation. Setúbal: SciTePress, 2010 - (Filipe, J.; Kacprzyk, J.), s. 316-321. ISBN 978-989-8425-32-4.
[ICNC 2010. International Conference on Neural Computation. Valencia (ES), 24.08.2010-26.08.2010]
R&D Projects: GA MŠMT OC10047
Institutional research plan: CEZ:AV0Z10300504
Keywords : learning from data * minimization of empirical and expected error functionals * reproducing kernel Hilbert spaces
Subject RIV: IN - Informatics, Computer Science

It is shown that application of methods from theory of inverse problems to learning from data leads to simple proofs of characterization of minima of empirical and expected error functionals and their regularized versions. The reformulation of learning in terms of inverse problems also enables comparison of regularized and non regularized case showing that regularization achieves stability by merely modifying output weights of global minima. Methods of theory of inverse problems lead to choice of reproducing kernel Hilbert spaces as suitable ambient function spaces.
Permanent Link: http://hdl.handle.net/11104/0189395