---
_id: '2008'
abstract:
- lang: eng
  text: "We concentrate our research activities on the multivariate feature selection,
    which is one important part of many machine learning tasks. In partucular, Linear
    Discriminant Analysis [1] belongs to the state-of-the-art methods for the multivariate
    analysis. From the theoretical point of view, it is the well-known fact that LDA
    is best suitable in the case the features are Gaussian distributed.\r\nIn the
    theoretical part of the presented paper, we analyse the properties of the multivariate
    discriminant analysis with respect to the feature selection. In this context,
    we consider a binary supervised learning task and assume that the features are
    Gaussian distributed. The discriminant analysis solves the mentioned supervised
    learning task by maximising of the discriminant value, calculated for the linear
    combination of the features.\r\nThe initial LDA solution a 2 Rd is considered
    for all given features from the feature space X \x1A Rd. The corresponding discriminant
    is calculated by the formula:\r\nd(a; x1, . . . , xd) := (μ+ − μ−)2\r\n\e2+\r\n+
    \e2−\r\n,\r\nwhere μ+/− are projected class means and \e2 +/− are projected class
    variances (with respect to a). We proof several propositions with the aim to find
    subsets of the features having higher discriminant value as original d(a; x1,
    . . . , xd). For the suitability in the real world settings, here we are interested
    in fast searching for such subsets.\r\nThe performance of the mentioned propositions
    is examined experimentally on datasets from UCI repository [2]. Several application
    scenarien will be discussed and tested on the datasets. In addition, tests show
    that the performance can be achieved also in the case the features are not Gaussian
    distributed."
author:
- first_name: Helene
  full_name: Dörksen, Helene
  id: '46416'
  last_name: Dörksen
- first_name: Volker
  full_name: Lohweg, Volker
  id: '1804'
  last_name: Lohweg
  orcid: 0000-0002-3325-7887
citation:
  ama: 'Dörksen H, Lohweg V. Multivariate Gaussian Feature Selection. . In: <i>European
    Conference on Data Analysis (ECDA2018)</i>. Paderborn, Germany; 2018.'
  apa: Dörksen, H., &#38; Lohweg, V. (2018). Multivariate Gaussian Feature Selection.
    . In <i>European Conference on Data Analysis (ECDA2018)</i>. Paderborn, Germany.
  bjps: <b>Dörksen H and Lohweg V</b> (2018) Multivariate Gaussian Feature Selection.
    . <i>European Conference on Data Analysis (ECDA2018)</i>. Paderborn, Germany.
  chicago: Dörksen, Helene, and Volker Lohweg. “Multivariate Gaussian Feature Selection.
    .” In <i>European Conference on Data Analysis (ECDA2018)</i>. Paderborn, Germany,
    2018.
  chicago-de: 'Dörksen, Helene und Volker Lohweg. 2018. Multivariate Gaussian Feature
    Selection. . In: <i>European Conference on Data Analysis (ECDA2018)</i>. Paderborn,
    Germany.'
  din1505-2-1: '<span style="font-variant:small-caps;">Dörksen, Helene</span> ; <span
    style="font-variant:small-caps;">Lohweg, Volker</span>: Multivariate Gaussian
    Feature Selection. . In: <i>European Conference on Data Analysis (ECDA2018)</i>.
    Paderborn, Germany, 2018'
  havard: 'H. Dörksen, V. Lohweg, Multivariate Gaussian Feature Selection. , in: European
    Conference on Data Analysis (ECDA2018), Paderborn, Germany, 2018.'
  ieee: H. Dörksen and V. Lohweg, “Multivariate Gaussian Feature Selection. ,” in
    <i>European Conference on Data Analysis (ECDA2018)</i>, Paderborn, 2018.
  mla: Dörksen, Helene, and Volker Lohweg. “Multivariate Gaussian Feature Selection.
    .” <i>European Conference on Data Analysis (ECDA2018)</i>, 2018.
  short: 'H. Dörksen, V. Lohweg, in: European Conference on Data Analysis (ECDA2018),
    Paderborn, Germany, 2018.'
  ufg: '<b>Dörksen, Helene/Lohweg, Volker (2018)</b>: Multivariate Gaussian Feature
    Selection. , in: <i>European Conference on Data Analysis (ECDA2018)</i>, Paderborn,
    Germany.'
  van: 'Dörksen H, Lohweg V. Multivariate Gaussian Feature Selection. . In: European
    Conference on Data Analysis (ECDA2018). Paderborn, Germany; 2018.'
conference:
  end_date: 2018-07-06
  location: Paderborn
  name: European Conference on Data Analysis
  start_date: 2018-07-04
date_created: 2019-11-25T08:35:48Z
date_updated: 2023-03-15T13:49:38Z
department:
- _id: DEP5023
keyword:
- multivariate feature selection
- Gaussian distribution
- linear discriminant analysis
language:
- iso: eng
place: Paderborn, Germany
publication: European Conference on Data Analysis (ECDA2018)
status: public
title: 'Multivariate Gaussian Feature Selection. '
type: conference
user_id: '15514'
year: 2018
...