A0324
Title: On decision making in cluster analysis
Authors: Christian Hennig - University of Bologna (Italy) [presenting]
Abstract: The purpose of the talk is to discuss some of the decisions that are required when using cluster analysis in practice. These may include connecting the aim of analysis to the choice of appropriate methodology, decisions at the preprocessing stage: standardization, transformation, dimension reduction, choice of distance measure, decisions regarding outliers and number of clusters, use of validation tools such as stability assessment, testing, and visualization, particularly for comparing different clusterings and deciding between them.
