Reducing the Training Times of Neural Classifiers with Dataset Condensing

  • Se-Ho Choi
  • Peter Rockett
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1876)


In this paper we apply a k-nearest-neighbour-based data condensing algorithm to the training sets of multi-layer perceptron neural networks. By removing the overlapping data and retaining only training exemplars adjacent to the decision boundary we are able to significantly speed the network training time while achieving an undegraded misclassification rate compared to a network trained on the unedited training set. We report results on a range of synthetic and real datasets which indicate that a speed-up of an order of magnitude in the network training time is typical.


Neural networks data editing pattern classifiers 


  1. [1]
    Richard M D, Lippmann R P, “Neural Network Classifiers Estimate Bayesian a Posterior Probabilities”, Neural Computation, Vol 3, pp 461–483, 1991.CrossRefGoogle Scholar
  2. [2]
    Kraaijveld M A, Duin R P W, “An Optimal Stopping Criterion for Backpropagation Learning”, Neural Network World, Vol 1, No 6, pp 365–370, June 1991.Google Scholar
  3. [3]
    Hara K, Nakayama K, “Training Data Selection Method for Generalization by Multilayer Neural Networks”, IEICE Trans Fundamentals, vol E81–A, No 3, pp 374–381, 1998.Google Scholar
  4. [4]
    Cachin C, “Pedagogical Pattern Selection Strategies”, Neural Networks, Vol 7, No 1, pp 175–181, 1994CrossRefGoogle Scholar
  5. [5]
    Leisch F, Jain L C, Hornik K, “Cross-Validation with Active Pattern Selection for Neural-Network Classifiers,” IEEE Trans on Neural Networks, vol 9, No 1, pp 35–41, 1998.CrossRefGoogle Scholar
  6. [6]
    Hwang J-N, Choi J J, Oh S, Marks R J, “Query-Based Learning Applied to Partially Trained Multilayer Perceptrons”, IEEE Trans Neural Networks, Vol 2, No 1, pp 131–136, 1991CrossRefGoogle Scholar
  7. [7]

Copyright information

© Springer-Verlag Berlin Heidelberg 2000

Authors and Affiliations

  • Se-Ho Choi
    • 1
  • Peter Rockett
    • 1
  1. 1.Department of Electronic and Electrical EngineeringUniversity of SheffieldSheffieldUK

Personalised recommendations