Jove
Visualize
Contact Us

Related Experiment Videos

Feedforward neural network with adaptive reference pattern layer.

M Lehtokangas1

  • 1Tampere University of Technology, Signal Processing Laboratory, Finland.

International Journal of Neural Systems
|July 13, 1999
PubMed
Summary

This study introduces a novel hybrid neural network combining Multilayer Perceptron (MLP) and Radial Basis Function (RBF) features. The new model demonstrates faster learning and a more compact structure compared to standard MLP and RBF networks.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Fast initialization for cascade-correlation learning.

IEEE transactions on neural networks·2008
Same author

Modified cascade-correlation learning for classification.

IEEE transactions on neural networks·2008
Same author

Modelling with constructive backpropagation.

Neural networks : the official journal of the International Neural Network Society·2003
Same author

Optimization of sigma amplitude threshold in sleep spindle detection.

Journal of sleep research·2001
Same author

Autoassociative MLP in sleep spindle detection.

Journal of medical systems·2000
Same author

Determining the number of centroids for CMLP network.

Neural networks : the official journal of the International Neural Network Society·2000
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Neural Networks

Background:

  • Standard Multilayer Perceptron (MLP) networks and Radial Basis Function (RBF) networks have limitations in learning efficiency and model complexity.
  • Developing hybrid models can potentially overcome the drawbacks of individual network architectures.

Purpose of the Study:

  • To investigate a novel hybrid neural network architecture combining MLP and RBF characteristics.
  • To evaluate the performance of the proposed hybrid model against standard MLP and RBF networks.

Main Methods:

  • A hybrid neural network architecture was developed, integrating an adaptive reference pattern layer into a standard MLP structure.
  • The adaptive reference pattern layer's units compute component-wise squared differences between reference patterns and input variables, similar to RBF hidden layers.

Related Experiment Videos

Main Results:

  • Benchmark experiments demonstrated that the proposed hybrid network achieves significantly faster and more efficient learning.
  • The hybrid model resulted in a more compact network structure compared to traditional MLP and RBF networks.

Conclusions:

  • The hybrid MLP-RBF neural network offers substantial advantages in learning speed and structural efficiency.
  • This novel architecture presents a promising alternative for various modeling applications requiring efficient and compact neural networks.