A machine learning method for generation of a neural network architecture: a continuous ID3 algorithm

Author

Cios, Krzysztof J. ; Liu, Ning

Author_Institution

Dept. of Electr. Eng., Toledo Univ., OH, USA

Volume

3

Issue

2

fYear

1992

fDate

3/1/1992 12:00:00 AM

Firstpage

280

Lastpage

291

Abstract

The relation between the decision trees generated by a machine learning algorithm and the hidden layers of a neural network is described. A continuous ID3 algorithm is proposed that converts decision trees into hidden layers. The algorithm allows self-generation of a feedforward neural network architecture. In addition, it allows interpretation of the knowledge embedded in the generated connections and weights. A fast simulated annealing strategy, known as Cauchy training, is incorporated into the algorithm to escape from local minima. The performance of the algorithm is analyzed on spiral data

Keywords

decision theory; entropy; learning systems; neural nets; simulated annealing; trees (mathematics); Cauchy training; architecture generation; continuous ID3 algorithm; decision trees; feedforward neural network architecture; hidden layers; machine learning method; self-generation; simulated annealing; spiral data; Algorithm design and analysis; Data analysis; Decision trees; Feedforward neural networks; Learning systems; Machine learning algorithms; Neural networks; Performance analysis; Simulated annealing; Spirals;

fLanguage

English

Journal_Title

Neural Networks, IEEE Transactions on

Publisher

ieee

ISSN

1045-9227

Type

jour

DOI

10.1109/72.125869

Filename

125869