Regularization and complexity control in feed-forward networks

Christopher M. Bishop

    Research output: Chapter in Book/Published conference output › Chapter

    Abstract

    In this paper we consider four alternative approaches to complexity control in feed-forward networks based respectively on architecture selection, regularization, early stopping, and training with noise. We show that there are close similarities between these approaches and we argue that, for most practical applications, the technique of regularization should be the method of choice.
    Original languageEnglish
    Title of host publicationProceedings International Conference on Artificial Neural Networks ICANN'95
    PublisherEC2 et Cie
    Pages141-148
    Number of pages8
    ISBN (Print)2-910085-19-8
    Publication statusUnpublished - 1995

    Bibliographical note

    International Conference on Artificial Neural Networks ICANN'95.

    Keywords

    • NCRG complexity control feed-forward networks architecture selection regularization early stopping training with noise

    Access to Document

    • NCRG_95_022.pdf

      International Conference on Artificial Neural Networks ICANN'95.

      Cite this