Estimating hidden units for two-layer perceptrons

M. Gutierrez, J. Wang, R. O. Grondin

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    6 Citations (Scopus)

    Abstract

    A method of estimating the number of hidden units required by a two-layer perception learning binary mappings using back-propagation of error signals is presented. A two-layer perception actually has three discernable layers. One is a layer of inputs, each of which is connected to every unit found on the next or hidden layer. These units are called the hidden units and each hidden unit, in turn, is connected to every output unit on the output layer. No intralayer connections are used. Usually the application dictates the number of input units and the number of output units in a rather obvious fashion. Specifying the number of hidden units however is more difficult and yet very important. We consider an example using a net that classifies 2004 input vectors. It would require only a set of 4 such vectors to present the net with a conflict similar to the exclusive-OR problem.

    Original languageEnglish (US)
    Title of host publicationIEE Conference Publication
    PublisherPubl by IEE
    Pages120-124
    Number of pages5
    Edition313
    StatePublished - 1989
    EventFirst IEE International Conference on Artificial Neural Networks - London, Engl
    Duration: Oct 16 1989Oct 18 1989

    Other

    OtherFirst IEE International Conference on Artificial Neural Networks
    CityLondon, Engl
    Period10/16/8910/18/89

    Fingerprint

    Neural networks
    Backpropagation

    ASJC Scopus subject areas

    • Electrical and Electronic Engineering

    Cite this

    Gutierrez, M., Wang, J., & Grondin, R. O. (1989). Estimating hidden units for two-layer perceptrons. In IEE Conference Publication (313 ed., pp. 120-124). Publ by IEE.

    Estimating hidden units for two-layer perceptrons. / Gutierrez, M.; Wang, J.; Grondin, R. O.

    IEE Conference Publication. 313. ed. Publ by IEE, 1989. p. 120-124.

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Gutierrez, M, Wang, J & Grondin, RO 1989, Estimating hidden units for two-layer perceptrons. in IEE Conference Publication. 313 edn, Publ by IEE, pp. 120-124, First IEE International Conference on Artificial Neural Networks, London, Engl, 10/16/89.
    Gutierrez M, Wang J, Grondin RO. Estimating hidden units for two-layer perceptrons. In IEE Conference Publication. 313 ed. Publ by IEE. 1989. p. 120-124
    Gutierrez, M. ; Wang, J. ; Grondin, R. O. / Estimating hidden units for two-layer perceptrons. IEE Conference Publication. 313. ed. Publ by IEE, 1989. pp. 120-124
    @inproceedings{aa7196fb599d4f45870fbda2263c712d,
    title = "Estimating hidden units for two-layer perceptrons",
    abstract = "A method of estimating the number of hidden units required by a two-layer perception learning binary mappings using back-propagation of error signals is presented. A two-layer perception actually has three discernable layers. One is a layer of inputs, each of which is connected to every unit found on the next or hidden layer. These units are called the hidden units and each hidden unit, in turn, is connected to every output unit on the output layer. No intralayer connections are used. Usually the application dictates the number of input units and the number of output units in a rather obvious fashion. Specifying the number of hidden units however is more difficult and yet very important. We consider an example using a net that classifies 2004 input vectors. It would require only a set of 4 such vectors to present the net with a conflict similar to the exclusive-OR problem.",
    author = "M. Gutierrez and J. Wang and Grondin, {R. O.}",
    year = "1989",
    language = "English (US)",
    pages = "120--124",
    booktitle = "IEE Conference Publication",
    publisher = "Publ by IEE",
    edition = "313",

    }

    TY - GEN

    T1 - Estimating hidden units for two-layer perceptrons

    AU - Gutierrez, M.

    AU - Wang, J.

    AU - Grondin, R. O.

    PY - 1989

    Y1 - 1989

    N2 - A method of estimating the number of hidden units required by a two-layer perception learning binary mappings using back-propagation of error signals is presented. A two-layer perception actually has three discernable layers. One is a layer of inputs, each of which is connected to every unit found on the next or hidden layer. These units are called the hidden units and each hidden unit, in turn, is connected to every output unit on the output layer. No intralayer connections are used. Usually the application dictates the number of input units and the number of output units in a rather obvious fashion. Specifying the number of hidden units however is more difficult and yet very important. We consider an example using a net that classifies 2004 input vectors. It would require only a set of 4 such vectors to present the net with a conflict similar to the exclusive-OR problem.

    AB - A method of estimating the number of hidden units required by a two-layer perception learning binary mappings using back-propagation of error signals is presented. A two-layer perception actually has three discernable layers. One is a layer of inputs, each of which is connected to every unit found on the next or hidden layer. These units are called the hidden units and each hidden unit, in turn, is connected to every output unit on the output layer. No intralayer connections are used. Usually the application dictates the number of input units and the number of output units in a rather obvious fashion. Specifying the number of hidden units however is more difficult and yet very important. We consider an example using a net that classifies 2004 input vectors. It would require only a set of 4 such vectors to present the net with a conflict similar to the exclusive-OR problem.

    UR - http://www.scopus.com/inward/record.url?scp=0024923296&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=0024923296&partnerID=8YFLogxK

    M3 - Conference contribution

    AN - SCOPUS:0024923296

    SP - 120

    EP - 124

    BT - IEE Conference Publication

    PB - Publ by IEE

    ER -