Closed sets for labeled data
Gemma C. Garriga, Petra Kralj and Nada Lavrac
In: 10th European Conference on Principles and Practice of Knowledge Discovery in Databases, 18 - 22 Sep 2006, Berlin, Germany.
Closed sets are being successfully applied in the context of
compacted data representation for association rule learning.
However, their use is mainly descriptive. In this paper we
will show that, when considering labeled data, closed sets
can be adapted for predictive purposes by conveniently
contrasting covering properties on positive and negative
examples. We formally justify that these sets characterize
the space of relevant combinations of features for discriminating
the target class. In practice, identifying relevant/irrelevant
combinations of features through closed sets is useful
in many applications. Here we apply it to emerging patterns,
subgroup discovery, and fast learning of relevant rules on
datasets characterized by highly unbalanced distribution.
|EPrint Type:||Conference or Workshop Item (Paper)|
|Project Keyword:||Project Keyword UNSPECIFIED|
|Subjects:||Theory & Algorithms|
|Deposited By:||Petra Kralj|
|Deposited On:||22 November 2006|