Logo des Repositoriums

Auflistung nach Autor:in "Kazempour, Daniyal"

1 - 2 von 2
  • Textdokument
    DICE: Density-based Interactive Clustering and Exploration
    (BTW 2019, 2019) Kazempour, Daniyal; Kazakov, Maksim; Kröger, Peer; Seidl, Thomas
    Clustering algorithms are mostly following the pipeline to provide input data, and hyperparameter values. Then the algorithms are executed and the output files are generated or visualized. We provide in our work an early prototype of an interactive density-based clustering tool named DICE in which the users can change the hyperparameter settings and immediately observe the resulting clusters. Further the users can browse through each of the single detected clusters and get statistics regarding as well as a convex hull profile for each cluster. Further DICE keeps track of the chosen settings, enabling the user to review which hyperparameter values have been previously chosen. DICE can not only be used in scientific context of analyzing data, but also in didactic settings in which students can learn in an exploratory fashion how a density-based clustering algorithm like e.g. DBSCAN behaves.
  • Zeitschriftenartikel
    On Methods and Measures for the Inspection of Arbitrarily Oriented Subspace Clusters
    (Datenbank-Spektrum: Vol. 21, No. 3, 2021) Kazempour, Daniyal; Winter, Johannes; Kröger, Peer; Seidl, Thomas
    When using arbitrarily oriented subspace clustering algorithms one obtains a partitioning of a given data set and for each partition its individual subspace. Since clustering is an unsupervised machine learning task, we may not have “ground truth” labels at our disposal or do not wish to rely on them. What is needed in such cases are internal measure which permits a label-less analysis of the obtained subspace clustering. In this work, we propose methods for revising clusters obtained from arbitrarily oriented correlation clustering algorithms. Initial experiments conducted reveal improvements in the clustering results compared to the original clustering outcome. Our proposed approach is simple and can be applied as a post-processing step on arbitrarily oriented correlation clusterings.