LEARNING-CURVES AND OPTIMIZATION OF A MULTILAYER PERCEPTRON NEURAL-NETWORK FOR CHROMOSOME CLASSIFICATION

B LERNER; H GUTERMAN; I DINSTEIN; Y ROMEM

LEARNING-CURVES AND OPTIMIZATION OF A MULTILAYER PERCEPTRON NEURAL-NETWORK FOR CHROMOSOME CLASSIFICATION

B LERNER, H GUTERMAN, I DINSTEIN, Y ROMEM

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

The use of multilayer perceptron (MLP) neural network (NN) as human chromosome classifier was studied. The MLP NN classifier was optimized in the sense of learning rate, momentum constant and training cycle, for the chromosome data. The MLP classifier learning curves were examined by measuring probability of correct test set classification for an increasing size of training sets. Only 10-20 examples were required for the MLP NN classifier to reach its ultimate performance disregarding the number of features used. To compare the results to relevant theory, we have calculated the entropic error (loss). The empirical dependence of the entropic error on the number of examples is highly comparable to 1/t function that is a learning curve.

Original language	English
Title of host publication	WORLD CONGRESS ON NEURAL NETWORKS-SAN DIEGO, - 1994 INTERNATIONAL NEURAL NETWORK SOCIETY ANNUAL MEETING, VOL 3
Pages	C248-C253
State	Published - 1994

Cite this

@inproceedings{19b7efbc095345308df25d48211f9a8d,

title = "LEARNING-CURVES AND OPTIMIZATION OF A MULTILAYER PERCEPTRON NEURAL-NETWORK FOR CHROMOSOME CLASSIFICATION",

abstract = "The use of multilayer perceptron (MLP) neural network (NN) as human chromosome classifier was studied. The MLP NN classifier was optimized in the sense of learning rate, momentum constant and training cycle, for the chromosome data. The MLP classifier learning curves were examined by measuring probability of correct test set classification for an increasing size of training sets. Only 10-20 examples were required for the MLP NN classifier to reach its ultimate performance disregarding the number of features used. To compare the results to relevant theory, we have calculated the entropic error (loss). The empirical dependence of the entropic error on the number of examples is highly comparable to 1/t function that is a learning curve.",

author = "B LERNER and H GUTERMAN and I DINSTEIN and Y ROMEM",

year = "1994",

language = "English",

isbn = "0-8058-1745-X",

pages = "C248--C253",

booktitle = "WORLD CONGRESS ON NEURAL NETWORKS-SAN DIEGO, - 1994 INTERNATIONAL NEURAL NETWORK SOCIETY ANNUAL MEETING, VOL 3",

}

LEARNING-CURVES AND OPTIMIZATION OF A MULTILAYER PERCEPTRON NEURAL-NETWORK FOR CHROMOSOME CLASSIFICATION. / LERNER, B ; GUTERMAN, H ; DINSTEIN, I et al.
WORLD CONGRESS ON NEURAL NETWORKS-SAN DIEGO, - 1994 INTERNATIONAL NEURAL NETWORK SOCIETY ANNUAL MEETING, VOL 3. 1994. p. C248-C253.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - LEARNING-CURVES AND OPTIMIZATION OF A MULTILAYER PERCEPTRON NEURAL-NETWORK FOR CHROMOSOME CLASSIFICATION

AU - LERNER, B

AU - GUTERMAN, H

AU - DINSTEIN, I

AU - ROMEM, Y

PY - 1994

Y1 - 1994

N2 - The use of multilayer perceptron (MLP) neural network (NN) as human chromosome classifier was studied. The MLP NN classifier was optimized in the sense of learning rate, momentum constant and training cycle, for the chromosome data. The MLP classifier learning curves were examined by measuring probability of correct test set classification for an increasing size of training sets. Only 10-20 examples were required for the MLP NN classifier to reach its ultimate performance disregarding the number of features used. To compare the results to relevant theory, we have calculated the entropic error (loss). The empirical dependence of the entropic error on the number of examples is highly comparable to 1/t function that is a learning curve.

AB - The use of multilayer perceptron (MLP) neural network (NN) as human chromosome classifier was studied. The MLP NN classifier was optimized in the sense of learning rate, momentum constant and training cycle, for the chromosome data. The MLP classifier learning curves were examined by measuring probability of correct test set classification for an increasing size of training sets. Only 10-20 examples were required for the MLP NN classifier to reach its ultimate performance disregarding the number of features used. To compare the results to relevant theory, we have calculated the entropic error (loss). The empirical dependence of the entropic error on the number of examples is highly comparable to 1/t function that is a learning curve.

M3 - Conference contribution

SN - 0-8058-1745-X

SP - C248-C253

BT - WORLD CONGRESS ON NEURAL NETWORKS-SAN DIEGO, - 1994 INTERNATIONAL NEURAL NETWORK SOCIETY ANNUAL MEETING, VOL 3

ER -

LEARNING-CURVES AND OPTIMIZATION OF A MULTILAYER PERCEPTRON NEURAL-NETWORK FOR CHROMOSOME CLASSIFICATION

Abstract

Fingerprint

Cite this