Quick Reference

1 In computer graphics, the collecting of nearby objects into groups so that their effect in a radiosity calculation with another well-separated group can be approximated by a single interaction.

2 An approach used in machine learning and data mining that attempts to place data into groups. The intention is that groupings or clusters will emerge that have meaning for the application problem (see concept learning). Clusters may be formed on the basis of some similarity features but their characteristics do not have to be predefined (in contrast to classifier systems).

Subjects: Computing.

Reference entries