Outliers influence to the point distance distribution normality within the data clusters

In order to verify the cluster analysis results, a normality test is being applied to the distribution of data point's distances from their cluster center. The presence of the outlier points within the input data can however influence this method in a negative way. Therefore, a normality test w...

Ausführliche Beschreibung

Bibliographische Detailangaben

Hauptverfasser:	Malkic, J., Sarajlic, N., Hadzic, D.
Format:	Tagungsbericht
Sprache:	eng
Schlagworte:	Algorithm design and analysis Cluster verification Clustering algorithms data mining distance distribution normality Gaussian distribution Histograms Shape Vectors
Online-Zugang:	Volltext bestellen
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!