๐”– Bobbio Scriptorium
โœฆ   LIBER   โœฆ

Discretizing continuous neural networks using a polarization learning rule

โœ Scribed by Lifeng Wang; H.D. Cheng

Publisher
Elsevier Science
Year
1997
Tongue
English
Weight
661 KB
Volume
30
Category
Article
ISSN
0031-3203

No coin nor oath required. For personal study only.

โœฆ Synopsis

Discrete neural networks are simpler than their continuous counterparts, can obtain more stable solutions, and their hidden layer representations are easier to interpret. This paper presents a polarization learning rule for discretizing multi-layer neural networks with continuous activation functions. This role forces the activation value of a neuron towards the two poles of its activation function. First, we use this role in the form of a modified error function to discretize the hidden units of a back-propagation network. Then, we apply the same principle to the second-order recurrent networks to solve grammatical inference problems. The experimental results are superior to the ones using existing approaches.

๐Ÿ“œ SIMILAR VOLUMES

A learning result for continuous-time re
A learning result for continuous-time recurrent neural networks
โœ Eduardo D. Sontag ๐Ÿ“‚ Article ๐Ÿ“… 1998 ๐Ÿ› Elsevier Science ๐ŸŒ English โš– 117 KB

The following learning problem is considered, for continuous-time recurrent neural networks having sigmoidal activation functions. Given a "black box" representing an unknown system, measurements of output derivatives are collected, for a set of randomly generated inputs, and a network is used to ap

Greedy rule generation from discrete dat
Greedy rule generation from discrete data and its use in neural network rule extraction
โœ Koichi Odajima; Yoichi Hayashi; Gong Tianxia; Rudy Setiono ๐Ÿ“‚ Article ๐Ÿ“… 2008 ๐Ÿ› Elsevier Science ๐ŸŒ English โš– 1018 KB

This paper proposes a GRG (Greedy Rule Generation) algorithm, a new method for generating classification rules from a data set with discrete attributes. The algorithm is "greedy" in the sense that at every iteration, it searches for the best rule to generate. The criteria for the best rule include t

Local unsupervised learning rules for a
Local unsupervised learning rules for a spiking neural network with dendrite
โœ Olivier FL Manette ๐Ÿ“‚ Article ๐Ÿ“… 2011 ๐Ÿ› BioMed Central ๐ŸŒ English โš– 316 KB

How could synapse number and position on a dendrite affect neuronal behavior with respect to the decoding of firing rate and temporal pattern? We developed a model of a neuron with a passive dendrite and found that dendritic length and the particular synapse positions directly determine the behavior