Impact of Mixed Metrics on Clustering

摘要

One of the features involved in clustering is the evaluation of distances between individuals. This paper is related with the use of mixed metrics for clustering messy data. Indeed, when facing complex real domains it becomes natural to deal simultaneously with numerical and symbolic attributes. This can be treated on different approaches. Here, the use of mixed metrics is followed. In the paper, a family of mixed metrics introduced by Gibert is used with different parameters on an experimental data set, in order to assess the impact on final classes.