Advertisement

Synthesizing Local Patterns by Weighting

  • Shichao Zhang
  • Chengqi Zhang
  • Xindong Wu
Chapter
  • 96 Downloads
Part of the Advanced Information and Knowledge Processing book series (AI&KP)

Abstract

For the identification of high-vote patterns and exceptional patterns, the projections of a pattern on subspaces were investigated in Chapters 7 and 8. Although each branch of a company, large or small, has equal power to vote for patterns as part of a general company decision, individual branches may be of different importance to their company. Also, local patterns may have different supports within a branch. To take these factors into account, this chapter presents a model for synthesizing global patterns from local patterns by weighting. This model synthesizes association rules with respect to all dimensions: name, vote, vsupp, and vconf.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Notes

  1. 1.
    It is not realistic to rediagnose the patient and aim for consistent results at each diagnosis step taken by the five experts. This is because some steps, such as operations, cannot be repeated, and, also, a rediagnosis can be very expensive.Google Scholar
  2. 2.
    This support differs from the support defined in Chapter 2. Here, the support is the number of databases that vote for the rule.Google Scholar

Copyright information

© Springer-Verlag London 2004

Authors and Affiliations

  • Shichao Zhang
    • 1
  • Chengqi Zhang
    • 1
  • Xindong Wu
    • 2
  1. 1.FIT, Universityof Technology SydneyAustralia
  2. 2.Department of Computer ScienceUniversity of VermontUSA

Personalised recommendations