HDBSCAN is an algorithm built on top of DBSCAN, by exploring values of its main hyperparameter (max distance for points to be considered neighbors). Hence, HDBSCAN builds clusters as areas of high density separated by areas of low density. Due to this rather generic view, clusters found by HDBSCAN can be any shape, as opposed to k-means which assumes that clusters are convex shaped. Numerical features should use standard rescaling.
Minimum ratio of train set records to form a cluster. Several ratio values will create several models. Ratio should be between 0 and 0.5