Evaluating the Performance of the DBSCAN Algorithm's using Number of External Scores Measures

Authors

Rajaa Hasan Abbas, University of Kufa, College of Education for Girls, Department of Mathematics, IraqFollow
Huda Qusay Hashim, AL-Furat AL-Awsat Technical University, Management Technical College, KufaFollow
Huda Karem Nasser, Department of Mathematics, IraqFollow

Abstract

The emergence of more informative clustering methods than classical representations is important, so the density-based spatial clustering for applications with noise (DBSCAN) technique can yield an accurate statistical idea of clusters. DBSCAN is becoming more and more popular. On the other hand, if we are aware of actual datasets so that we can make comparisons with these datasets, we aim to determine the accuracy with which the partitioning is estimated using the density-based method. Therefore, in order to compare the success of the partitioning found by the density-based approach under different models, some external scores measures (Adjusted Rand, F-measure, and Jaccard) are chosen in this paper to evaluate this method

Recommended Citation

Abbas, Rajaa Hasan; Hashim, Huda Qusay; and Nasser, Huda Karem (2025) "Evaluating the Performance of the DBSCAN Algorithm's using Number of External Scores Measures," Iraqi Journal for Computer Science and Mathematics: Vol. 6: Iss. 3, Article 41.
DOI: https://doi.org/10.52866/2788-7421.1310
Available at: https://ijcsm.researchcommons.org/ijcsm/vol6/iss3/41