An efficient paradigm for achieving synchronization in hierarchical clustering

Clustering is one of the important streams in data mining which is useful for discovering groups and identifying interesting distributions in the underlying data. For efficient clustering process in high dimensional datasets, synchronization algorithm is proposed. It can perform clustering by using the mathematical model called kuramato model. This model is widely used for the synchronization of oscillators. We regard each object in the dataset as phase oscillator and do the clustering process. This algorithm doesn´t requires any input parameters which are difficult to estimate. It only requires the efficient interaction range as input for the clustering process. Since it is a dynamic clustering algorithm it does not requires the users to determine the predefined inputs like the number of clusters to be formed. Our proposed algorithm is scalable even when the dimensionality of the dataset increases. Hierarchical clustering algorithm can generate hierarchical tree like clusters which is used to predict the clusters of different types in high dimensional real time datasets.