Review of Dynamic Means Partition based Clustering algorithm for Time Series Data

In this paper, a review of dynamic means partition based clustering algorithm for time series data is purposed in which a set of unlabeled time-series are partitioned into groups or clusters where all the objects grouped in the same cluster should be coherent or homogeneous. The proposed system is a prediction system to predict the gold market rates based on the history data set. These systems divide the data set in small fragments and process on each data block separately and conclude them as a single unit. Finally a layered neural network is implemented to filter the data and to derive the conclusion.