Post-processing Two post-processing strategies that decrease the number of clusters, while trying to minimize the increase in total Sse, are Disperse a cluster This is accomplished by removing the centroid that corresponds to the cluster. The points in that cluster are then re-assigned to other clusters The cluster that is dispersed should be the one that increases the total sse the least Merge two clusters We can merge the two clusters that result in the smallest increase in total SSE 8/20/2016 PATTERN RECOGNITIONPost-processing Two post-processing strategies that decrease the number of clusters, while trying to minimize the increase in total SSE, are ◦ Disperse a cluster ◦ This is accomplished by removing the centroid that corresponds to the cluster. ◦ The points in that cluster are then re-assigned to other clusters. ◦ The cluster that is dispersed should be the one that increases the total SSE the least. ◦ Merge two clusters ◦ We can merge the two clusters that result in the smallest increase in total SSE. 8/20/2016 PATTERN RECOGNITION 32