Factor PD-Clustering

Factorial clustering methods have been developed in recent years thanks to the improving of computational power. These methods perform a linear transformation of data and a clustering on transformed data optimizing a common criterion. Factor PD-clustering is based on Probabilistic Distance clustering (PD-clustering). PD-clustering is an iterative, distribution free, probabilistic, clustering method. Factor PD-clustering makes a linear transformation of original variables into a reduced number of orthogonal ones using a common criterion with PD-Clustering. This paper demonstrates that Tucker3 decomposition permits to obtain this transformation. Factor PDclustering exploits alternatively Tucker3 decomposition and PD-clustering on transformed data until convergence is achieved. This method can significantly improve the algorithm performance; large datasets can thus be partitioned into clusters with increasing stability and robustness of the results.