Placing spline knots in neural networks using splines as activation functions

Hlavackova, Katerina;Verleysen, Michel
(1997) Neurocomputing — Vol. 17, n° 3-4, p. 159-167 (1997)

Files

pdfdocument.pdf
  • Restricted Access
  • Adobe PDF
  • 455.72 KB
Request a copy

Details

Authors
  • Hlavackova, KaterinaAcademy of Sciences of the Czech Republic (Prague)
    Author
  • Verleysen, MichelUCLouvain
    Author
Abstract
When using feed-forward neural networks with spline activation functions, the quality of approximation depends on the knot placement of spline functions. We demonstrate a method of choosing equidistant knots in each subdivision of the space when an arbitrary initial division is given, in order to keep the approximation error under a predefined limit.
Affiliations

Citations

  • APA
  • Chicago
  • FWB

Hlavackova, K., & Verleysen, M. (1997). Placing spline knots in neural networks using splines as activation functions. Neurocomputing, 17(3-4), 159-167. https://doi.org/10.1016/S0925-2312(97)00053-2 (Original work published 1997)