Please use this identifier to cite or link to this item:
|Title:||Model selection in function approximation problem based on statistical learning theory|
|Authors:||Burnaev, E. V.|
Belyaev, M. G.
|Keywords:||ЭБ БГУ::ОБЩЕСТВЕННЫЕ НАУКИ::Информатика|
|Abstract:||In order to approximate multidimensional function it is necessary to select the complexity of the model used for approximation of unknown function. Method on basis of statistical learning theory for complexity estimation of a model is elaborated. Proposed method is significantly less computationally intensive compared to such classical methods as AIC or cross-validation.|
|Appears in Collections:||Section 8. COMPUTER DATA ANALYSIS IN APPLICATIONS|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.