Short Communication on QUIST: A Quick Clustering Algorithm

Then, it starts splitting C into smaller sub-clusters, until either one of the following conditions holds, whichever happens first: 1. The instances within each sub-cluster, c, are too similar to be divided any further, or 2. The number of clusters hits an optional upper bound provided by the user. If such bound is not provided by the user, it is assume to be equal to the number of input instances, C Additionally, splitting a given cluster stops when it reaches a minimum cluster size per the user's choice. To decide whether or not instances within a cluster, c, are similar enough to stop splitting it, QUIST calculates the "spreadness" metric denoted by Ψ, such that the spreadness of c, denoted by Ψ