Evaluating Production Companies on Borsa Istanbul In Terms of Financial Indicators: Clustering Companies Using K-Means with the Silhouette Index and Elbow Method

The main aim of this study is to cluster production companies in Borsa Istanbul based on financial ratios using the k-means clustering method and to determine the natural structure of these companies. The analysis focuses on 11 years of production companies based on 15 financial ratios and two finan...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Muhasebe enstitüsü dergisi (Online) číslo 69; s. 1 - 19
Hlavní autori: Çiğdem Özarı, Esin Nesrin Can
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Istanbul University Press 01.08.2023
Predmet:
elbow method
financial indicators
k-means clustering method
silhouette index
ISSN:2667-6982
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:The main aim of this study is to cluster production companies in Borsa Istanbul based on financial ratios using the k-means clustering method and to determine the natural structure of these companies. The analysis focuses on 11 years of production companies based on 15 financial ratios and two financial indicators. The study determined the similarity among the financial performances of these companies using the k-means clustering analysis and evaluated the most appropriate cluster group and the most appropriate number of clusters that should be separated using the silhouette index and elbow method. The analysis was performed with different initial centers within the dataset for each k value by considering the choice of the initial center, which is a drawback of the k-means clustering method. In addition, because the number of clusters in which production companies should be grouped naturally is unknown, different k values were examined, with the most appropriate number of clusters to be separated from these values being determined using the silhouette index and elbow method. The results of the clustering analysis imply that splitting production companies into two clusters provides more accurate results.
ISSN:2667-6982
DOI:10.26650/MED.1278850