论文信息 - An extended procedure of constructing neural networks for supervised dichotomy

An extended procedure of constructing neural networks for supervised dichotomy

K.J. Clos and N. Liu (1992) proposed a neural network generation procedure for supervised two-class discretization based on a continuous ID3 algorithm. The method constructs neural networks consisting of neurons with linear activation functions. We extend the procedure to allow decision boundaries to be any arbitrary function. The advantage of the extension is revealed by the decrease in size of the generated neural networks.

Shie-Jue Lee | Mu-Tune Jone

[1] Griff L. Bilbro,et al. Maximum Entropy and Learning Theory , 1992, Neural Computation.

[2] Krzysztof J. Cios,et al. A machine learning method for generation of a neural network architecture: a continuous ID3 algorithm , 1992, IEEE Trans. Neural Networks.

[3] S. Y. Kung,et al. An algebraic projection analysis for optimal hidden units size and learning rates in back-propagation learning , 1988, IEEE 1988 International Conference on Neural Networks.

[4] Sun-Yuan Kung,et al. Digital neural networks , 1993, Prentice Hall Information and System Sciences Series.

[5] Peter Seitz,et al. Minimum class entropy: A maximum information approach to layered networks , 1989, Neural Networks.

[6] Hsin-Chia Fu,et al. On the Capability of a New Multilayer Perceptron for Two-Class Classification Problems , 1992, J. Inf. Sci. Eng..

[7] H. Szu. Fast simulated annealing , 1987 .

[8] F ROSENBLATT,et al. The perceptron: a probabilistic model for information storage and organization in the brain. , 1958, Psychological review.

[9] I. Bratko,et al. Information-based evaluation criterion for classifier's performance , 2004, Machine Learning.

[10] K. Lang,et al. Learning to tell two spirals apart , 1988 .

[11] L. F. A. Wessels,et al. The Physical Correlates of Local Minima , 1990 .