Comparison of Distance and Linkage in Integrated Cluster Analysis with Multiple Discriminant Analysis on Home Ownership Credit Bank in Indonesia
Year: 2021
Type: article
Abstract: This study aims to compare the size of distance (Euclidean distance, Manhattan distance, and Mahalanobis distance) and linkage (average linkage, single linkage, and complete linkage) in integrated cluster analysis with Multiple Discriminant Analysis on Home Ownership Credit Bank consumers in Indonesia. The data used are secondary data from the 5C assessment on Bank consumers in Indonesia. The data contain notes on the 5 C assessment as well as 3 credit collectability (current, special mention, and substandard) from Home Ownership Credit customers. The population in this study were all Home Ownership Credit customers in all banks in Indonesia. The sampling technique used was purposive random sampling. The sample size is 300 customers from customer data at three branches of Bank in Indonesia. This research is a quantitative study using cluster analysis integrated with multiple discriminant analysis. The best method for classifying Home Ownership Credit Bank customers based on the 5C variable assessment is an integrated cluster analysis with Multiple Discriminant Analysis based on the Mahalanobis distance with 2 clusters, namely the high cluster and the low cluster. Use of an integrated cluster with Multiple Discriminant Analysis to compare distance and linkage measures. In addition, the objects used are Home Ownership Credit Bank customers in Indonesia.
Source: Mathematics and Statistics
Institution: University of Brawijaya
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4200433872