Density Estimation as a Preprocessing Step for Constructive Algorithms

Jörg C. Lemm, Valeriu Beiu, John G. Taylor

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

    Abstract

    The aim of the paper is to combine the generalization power of classical density estimation techniques with the efficiency of VLSI-friendly, constructive algorithms. This is important for VLSI-implementations of classification networks in the case of noisy data to avoid the well-known “overfitting” effects. The method consists of three steps 1) estimation of the probability densities for each class, 2) discretization of the input space and 3) describing the resulting classification regions using only easy implementable boolean AND and OR gates and comparisons. The “noisy spirals” classification problem, a noisy variant of the “two spirals” benchmark, is used for demonstration.
    Original languageEnglish
    Title of host publicationProceedings of the Third Annual SNN Symposium on Neural Networks
    DOIs
    Publication statusPublished - Sep 14 1995
    EventSNN'95 - Nijmegen, The Netherlands
    Duration: Sep 14 1995 → …

    Conference

    ConferenceSNN'95
    Period9/14/95 → …

    Fingerprint

    Dive into the research topics of 'Density Estimation as a Preprocessing Step for Constructive Algorithms'. Together they form a unique fingerprint.

    Cite this