|
Abstract:
|
Scaling of multivariate data (to put all variables on an "equal footing") prior to cluster analysis is important as a preprocessing step. Widely-used current methods for this are problematic.In this paper, we propose some intuitive alternatives, which we call "equalizers", that are directed at avoiding a basic pitfall in the usual methods.In addition, we let the data suggest weights or "highlighters", that emphasize those variables with most promise for revealing the latent cluster structure. The equalizing and highlighting schemes vary in degree of complexity from very simple weights based on order statistics to more complicated iterative ones. The methods are applied to simulated and real data sets, and their performances are compared to some currently used methods. The results indicate that, in many situations, the new methods are much better than the most popular method, autoscaling.
|
