Comparison of Stability of Classical Taxonomy Bagging Metod with Bagging Based on Co-Occurence Data
Ensemble approach has been successfully applied in the context of supervised
learning to increase the accuracy and stability of classification. Recently, analogous techniques for
cluster analysis have been suggested in order to increase classification accuracy, robustness and
stability of the clustering solutions. Research has proved that, by combining a collection of different
clusterings, an improved solution can be obtained.
The stability of a clustering algorithm with respect to small perturbations of data (e.g., data
subsampling or small variations in the feature values) or the parameters of the algorithm (e.g.,
random initialization) is a desirable quality of the algorithm. On the other hand, ensembles benefit
from diverse clusterers. Although built upon unstable components, the ensemble is expected to be
more accurate and robust than the individual clustering method. Here, we look at the stability of
the ensemble methods based on bagging idea and co-occurrence matrix. This paper carries out an
experimental study to compare stability of bagging method used to the classical data set with
bagging based on co-occurrence matrix.