Peter Brezany
      University of Vienna, Institute of Scientific Computing

              "Machine Learning and the Grid"

Machine learning is concerned with developing methods for software to
learn from experience or extract knowledge from examples in a database.
One important machine learning task is classification. This talk
addresses both classification methods developed for
traditional data mining in relational databases and text mining.

Classification deals with discovery of a predictive learning function
that classifies a data object into one of several predefined classes.
We present novel decision-tree-based classification services
which can be used either autonomously or as a building block
to construct distributed and scalable classifiers
that operate on data repositories integrated into the Grid that
typically involve large, complex, heterogeneous,
and geographically distributed datasets. Scalability and performance of 
the
prototype implementations within our GridMiner project (www.gridminer.org)
have been examined and the results will presented in the talk.
I will also provide an on-line demonstration of the classifiers 
functionality
as it is embedded into the GridMiner infrastructure.