Learning from Data Streams: an overview. Invited talk. José AGUILAR-RUIZ
La revue MODULAD, numéro 36, Juillet 2007
Classification is a very well-studied
task in data mining. In the last years, important works have been
published to scale up classification
algorithms in order to handle
large datasets. However, due to the high rate of streams of data,
a number of emerging
applications are demanding new approaches. Rule learning is an
efficient alternative to address non-stationary environments.
The talk presents an overview of rule-based learning
algorithms for data streams and emphasizes some important aspects
of these techniques.
Keywords: Data Streams, Rule-based learning.
Download paper : Rule
Learning from Data Streams: an overview