Note to users. If you're seeing this message, it means that your browser cannot find this page's style/presentation instructions -- or possibly that you are using a browser that does not support current Web standards. Find out more about why this message is appearing, and what you can do to make your experience of our site the best it can be.
Active Motif, Inc.

Site Tools

  • AAAS
  • Subscribe
  • Feedback

Site Search

Search Advanced

Science 16 February 2007:
Vol. 315. no. 5814, p. 909
DOI: 10.1126/science.315.5814.909d
Prev | Table of Contents | Next

This Week in Science

Figure 1 Complex data sets can be more readily analyzed if representative examples can be identified. Such "exemplars" might be points around which data will cluster, archetypal faces among a gallery of actual photos, or possible exons in a gene sequence. Unfortunately, extracting exemplars is computationally intensive, and conventional techniques only work well with numerical measures of data-point similarity and if the initial guess is close. Frey and Dueck (p. 972, published online 11 January; see the Perspective by Mézard) now report a method that enables much faster exemplar detection. The algorithm works by having the data points exchange "messages" that communicate whether a particular point could be an exemplar; iteration of the message-passing process allows dramatically faster processing as certain data points emerge as truly representative.

CREDIT: FREY AND DUECK




ADVERTISEMENT
Click Me!

ADVERTISEMENT
Click Me!

To Advertise     Find Products

Science. ISSN 0036-8075 (print), 1095-9203 (online)