On the use of simple classifiers for the initialisation of one-hidden-layer neural nets
January 1995
Research Paper
| Related Files |
|---|
View PDF Version
(eur-few-cs-95-08.pdf, 0.2MB) |
In this report we discuss the use of two simple classifiers to initialise the input-to-hidden layer of a one-hidden-layer neural network. These classifiers divide the input space in convex regions that can be represented by membership functions. These functions are then used to determine the weights of the first layer of a feedforward network.
Keywords
- LVQ-networks
- simple perceptrons
- mapping decision trees onto neural networks
- initialisation of feedforward networks
Automatically Extracted Terms
- function
- membership functions
- decision
- region
- decision tree
- membership
- mapping
- figure
- layer
- constant
- network
- hyperplane
- theorem
- polynomial
- input space
- space
- section
- input
- example
- constants c j
