Clustering centroid finding algorithm (CCFA) using spatial temporal data mining concept

Abstract

The main aim of the research focuses the clustering centroid value for spatio-temporal data mining. Using k-means, advanced k-means algorithm and Avg Centroid (AC) clustering. The real time data of the hurricane Indian Ocean 2001 to 2010 maximum wind details are focused in this paper. The clustering is taking as selection window method, the first window form the basis of the pixel coordinate value of the screen, the second clustering window one half of the centre point value. The data mining retrieves clustering data form basis of the selection window. Here to discuss k-means algorithmic steps are very few and same iteration is continuing till the same to get the centroid point. The enhanced k-means algorithm taken more steps but result is accurate algorithmic finishing stage; iteration also repeated very minimum times. The final discussion of this paper collects average centroid clustering for all previously selected values and current selected clustering data. The result of this paper gave the comparative study of the k-means, enhanced k-means algorithms and AC clustering values.