• DocumentCode
    1819568
  • Title

    A polynomial time algorithm for generating neural networks for classification problems

  • Author

    Roy, Asim ; Mukhopadhyay, Somnath

  • Author_Institution
    Dept. of Decision & Inf. Syst., Arizona State Univ., Tempe, AZ, USA
  • Volume
    1
  • fYear
    1992
  • fDate
    7-11 Jun 1992
  • Firstpage
    147
  • Abstract
    A novel polynomial time algorithm for the construction and training of multilayer perceptrons for classification problems is presented. It uses linear programming models to generate incrementally the hidden layer in a restricted higher-order perceptron. The polynomial time complexity of the method is proven and computational results are provided for some well-known problems. In all cases, very small nets were created compared to those reported previously
  • Keywords
    computational complexity; feedforward neural nets; linear programming; hidden layer; linear programming models; multilayer perceptrons; polynomial time algorithm; polynomial time complexity; restricted higher-order perceptron; Classification algorithms; Information systems; Linear programming; Multi-layer neural network; Multilayer perceptrons; Neural networks; Polynomials; Shape; Supervised learning; Testing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Neural Networks, 1992. IJCNN., International Joint Conference on
  • Conference_Location
    Baltimore, MD
  • Print_ISBN
    0-7803-0559-0
  • Type

    conf

  • DOI
    10.1109/IJCNN.1992.287225
  • Filename
    287225