Discretization of Target Attributes for Subgroup Discovery

Abstract. We describe an algorithm called TargetCluster for the discretization of continuous targets in subgroup discovery. The algorithm identifies patterns in the target data and uses them to select the discretization cutpoints. The algorithm has been implemented in a subgroup discovery method. Tests show that the discretization method likely leads to improved insight. Key words: Subgroup Discovery, Logic, Classification, Feature Selection

• Type: Conference
• Year: 2009
• Where: MLDM
• Authors: Katherine Moreland, Klaus Truemper
• Keywords: Algorithm, Discretization, Machine Learning, MLDM 2009, Subgroup Discovery