A novel modification to backpropagation sample selection strategy

László Rédei, Hans Wallinga

    Research output: Contribution to journalArticleAcademicpeer-review

    1 Citation (Scopus)
    53 Downloads (Pure)

    Abstract

    Random sample selection method in backpropagation results in convergence on the error (root of mean squared error, RMSE) surface. These problems, which are caused by the extreme (worst-case) errors, can be solved by a different sample selection strategy. A sample selection strategy has been proposed, which provides lower maximal errors and a higher confidence level on the expense of slightly increased RMSE. Applications are presented in the field of spectroscopic ellipsometry (SE), a sensitive, non-destructive but indirect analytical technique. Demonstrative example shows feature common to simulated annealing in the sense of escaping local minima.
    Original languageEnglish
    Pages (from-to)233-236
    JournalNuclear instruments & methods in physics research. Section A : Accelerators, spectrometers, detectors and associated equipment
    Volume389
    Issue number1-2
    DOIs
    Publication statusPublished - 1997
    EventNew Computing Techniques in Physics Research V - AIHENP International Workshop 1996 - Lausanne, Switzerland
    Duration: 2 Sep 19966 Sep 1996
    Conference number: 5

    Fingerprint Dive into the research topics of 'A novel modification to backpropagation sample selection strategy'. Together they form a unique fingerprint.

    Cite this