Figure 5

Kneeplot of the mean average dissimilarity across varying thresholds for the leader clustering algorithm. The mean average dissimilarity is computed as $\frac{1}{K}{\displaystyle {\sum }_{k=1}^K\frac{1}{\left|C_k\right|}}{\displaystyle {\sum }_{i\in C_k}^{}{\displaystyle {\sum }_{i\text{'}\in C_k}^{}{d}_{i,i\text{'}}}}$, with C _k being the data points in cluster k, and K being the total number of clusters. As the result of the leader algorithm depends on the input order of the data, the clustering procedure is repeated ten times with shuffled input.