Application of dynamic clustering using ADE to transportation planning

Abstract

In recent past, tremendous work has been done to find optimal number of clusters at run time for partitional clustering algorithms. Various Evolutionary Computation techniques have been used by researchers to evolve most appropriate number of clusters for different clustering problems. In this paper, we attempt to apply a new variant of adaptive differential evolution technique on a real world data set to find optimal number of clusters at runtime. The DCADE algorithm has been applied on Home Interview Survey (HIS) data related to a Transportation Project. Later clusters are formed and analyzed which are in accordance with the domain expert. © 2012 Springer-Verlag GmbH Berlin Heidelberg.