Constructive Feedforward Neural Networks for Regression Problems: a Survey

  title={Constructive Feedforward Neural Networks for Regression Problems: a Survey},
  author={SurveyTin-Yau Kwok and Dit-Yan Yeung},
In this paper, we review the procedures for constructing feedforward neural networks in regression problems. While standard back-propagation performs gradient descent only in the weight space of a network with xed topology, constructive procedures start with a small network and then grow additional hidden units and weights until a satisfactory solution is found. The constructive procedures are categorized according to the resultant network architecture and the learning algorithm for the network… CONTINUE READING
23 Citations
22 References
Similar Papers


Publications referenced by this paper.
Showing 1-10 of 22 references

Parallelizing the cascadecorrelation algorithm using TimeWarp

  • J. W. Shavlik
  • Neural Networks
  • 1995

Ridge polynomial networks

  • W. C. Schmidt.
  • IEEE Transactions on NeuralNetworks
  • 1995

Use of a quasiNewton method in a feedforward neuralnetwork construction algorithm

  • J. Ghosh.
  • IEEE Transactions on Neural Networks
  • 1995

Automatic smoothing spline projection pursuit

  • C. L. Wu A. Saha, D. S. Tang
  • Journalof Computational and Graphical Statistics
  • 1994

Connectionist projection pursuit regression

  • D. E. Rumelhart A. S. Weigend, B. A. Huberman.
  • Computa - tional Economics
  • 1994

Neural networks and evoluntionary computation . Part 1 : Hybrid approachesin arti cial intelligence

  • G. Weiss.
  • Proceedings of the First IEEE Conference on…
  • 1994

Neural networks and related methods for classi cation ( with discussion )

  • B. D. Ripley.
  • Journal of the Royal Statistical Society Series B
  • 1994

Similar Papers

Loading similar papers…