---
OA_type: closed access
_id: '3700'
abstract:
- lang: eng
  text: We propose a new method to partition an unlabeled dataset, called Discriminative
    Context Partitioning (DCP). It is motivated by the idea of splitting the dataset
    based only on how well the resulting parts can be separated from a context class
    of disjoint data points. This is in contrast to typical clustering techniques
    like K-means that are based on a generative model by implicitly or explicitly
    searching for modes in the distribution of samples. The discriminative criterion
    in DCP avoids the problems that density based methods have when the a priori assumption
    of multimodality is violated, when the number of samples becomes small in relation
    to the dimensionality of the feature space, or if the cluster sizes are strongly
    unbalanced. We formulate DCP‘s separation
    property as a large-margin criterion, and show how the resulting optimization
    problem can be solved efficiently. Experiments on the MNIST and USPS datasets
    of handwritten digits and on a subset of the Caltech256 dataset show that, given
    a suitable context, DCP can achieve good results even in situation where density-based
    clustering techniques fail.
acknowledgement: This work was funded in part by the EC project CLASS, IST 027978.
article_processing_charge: No
author:
- first_name: Christoph
  full_name: Lampert, Christoph
  id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
  last_name: Lampert
  orcid: 0000-0001-8622-7887
citation:
  ama: 'Lampert C. Partitioning of image datasets using discriminative context information.
    In: IEEE; 2008:1-8. doi:<a href="https://doi.org/10.1109/CVPR.2008.4587448">10.1109/CVPR.2008.4587448</a>'
  apa: 'Lampert, C. (2008). Partitioning of image datasets using discriminative context
    information (pp. 1–8). Presented at the CVPR: Computer Vision and Pattern Recognition,
    IEEE. <a href="https://doi.org/10.1109/CVPR.2008.4587448">https://doi.org/10.1109/CVPR.2008.4587448</a>'
  chicago: Lampert, Christoph. “Partitioning of Image Datasets Using Discriminative
    Context Information,” 1–8. IEEE, 2008. <a href="https://doi.org/10.1109/CVPR.2008.4587448">https://doi.org/10.1109/CVPR.2008.4587448</a>.
  ieee: 'C. Lampert, “Partitioning of image datasets using discriminative context
    information,” presented at the CVPR: Computer Vision and Pattern Recognition,
    2008, pp. 1–8.'
  ista: 'Lampert C. 2008. Partitioning of image datasets using discriminative context
    information. CVPR: Computer Vision and Pattern Recognition, 1–8.'
  mla: Lampert, Christoph. <i>Partitioning of Image Datasets Using Discriminative
    Context Information</i>. IEEE, 2008, pp. 1–8, doi:<a href="https://doi.org/10.1109/CVPR.2008.4587448">10.1109/CVPR.2008.4587448</a>.
  short: C. Lampert, in:, IEEE, 2008, pp. 1–8.
conference:
  name: 'CVPR: Computer Vision and Pattern Recognition'
date_created: 2018-12-11T12:04:41Z
date_published: 2008-09-18T00:00:00Z
date_updated: 2026-05-28T10:19:40Z
day: '18'
doi: 10.1109/CVPR.2008.4587448
extern: '1'
language:
- iso: eng
main_file_link:
- url: http://pub.ist.ac.at/~chl/papers/lampert-cvpr2008b.pdf
month: '09'
oa_version: None
page: 1 - 8
publication_status: published
publisher: IEEE
publist_id: '2657'
status: public
title: Partitioning of image datasets using discriminative context information
type: conference
user_id: 317138e5-6ab7-11ef-aa6d-ffef3953e345
year: '2008'
...