*5.2. Results*

Similarly to the experimental evaluation performed on state-of-art datasets, we first assessed the best parameter settings for a fair comparison between CHD, DBSCAN, HDB-SCAN, and OPTICS-Xi. We run several experimental tests to find the parameter settings capable of detecting the highest-quality city hotspots in terms of significance, compactness, and separability. Table 2 shows, for each algorithm, the selected input parameters and some statistics related to the achieved results. In particular, for each algorithm, the table reports the input parameter setting, the number of detected hotspots, the percentage of noise points, and the achieved Silhouette index values. In particular, Silhouette is an internal criterion to compute and evaluate clustering quality, and it is a measure of how similar an object is to its own cluster (cohesion) compared to other clusters (separation). The silhouette ranges from −1 to +1, where high values indicate that instances are well-matched to their own cluster and poorly matched to neighboring clusters. Thus, the higher the Silhouette value, the better the clustering quality (a more detailed description of this metric is reported in [33]). The hotspots detected by the considered algorithms are depicted in Figure 11, where they are highlighted through different colors, while noise points are black-colored.

**Table 2.** Overview of the results obtained by CHD, DBSCAN, HDBSCAN, and OPTICS-Xi.


**Figure 11.** The *Crime* dataset: detected clusters. (**a**) CHD. (**b**) DBSCAN. (**c**) HDBSCAN. (**d**) OPTICS-Xi.

Now, by observing the hotspots detected by the algorithms and shown in Figure 11, and the values reported in Table 2, we can make some considerations:


