Monthly
288 pp. per issue
6 x 9, illustrated
ISSN
0899-7667
E-ISSN
1530-888X
2014 Impact factor:
2.21

Neural Computation

Summer 1990, Vol. 2, No. 2, Pages 198-209
(doi: 10.1162/neco.1990.2.2.198)
© 1990 Massachusetts Institute of Technology
The Upstart Algorithm: A Method for Constructing and Training Feedforward Neural Networks
Article PDF (551.7 KB)
Abstract

A general method for building and training multilayer perceptrons composed of linear threshold units is proposed. A simple recursive rule is used to build the structure of the network by adding units as they are needed, while a modified perceptron algorithm is used to learn the connection strengths. Convergence to zero errors is guaranteed for any boolean classification on patterns of binary variables. Simulations suggest that this method is efficient in terms of the numbers of units constructed, and the networks it builds can generalize over patterns not in the training set.