Title:
|
Neural networks with periodic and monotonic activation functions: a comparative study in classification problems
|
Author:
|
Romero Merino, Enrique; Sopena, Josep Maria; Alquézar Mancho, René; Moliner, Joan L.
|
Other authors:
|
Universitat Politècnica de Catalunya. Departament de Llenguatges i Sistemes Informàtics; Universitat Politècnica de Catalunya. SOCO - Soft Computing |
Abstract:
|
This article discusses a number of reasons why the use of
non-monotonic functions as activation functions can lead to a marked
improvement in the performance of a neural network. Using a wide range
of benchmarks we show that a multilayer feed-forward network using
sine activation functions (and an appropriate choice of initial
parameters) learns much faster than one incorporating sigmoid
functions - as much as 150-500 times faster - when both types are
trained with backpropagation. Learning speed also compares favorably
with speeds reported using modified versions of the backpropagation
algorithm. In addition, computational and generalization capacity
increases. |
Subject(s):
|
-Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial -Non-monotonic functions -Neural networks |
Rights:
|
|
Document type:
|
Article - Published version Report |
Share:
|
|