Distance Analysis Measuring for Clustering using K-Means and Davies Bouldin Index Algorithm
Distance Analysis Measuring for Clustering using K-Means and Davies Bouldin Index Algorithm

Summary/Abstract: The purpose of this research is to analyze mapping results in the form of clusters formed using clustering method measures. This is done to determine the connections that the existing clusters create. Some of the measurements used are mixed measurements, Bregman differences, and number measurements (Mixed Euclidean Distance, Generalized Divergence, Squared Euclidean Distance, Mahalanobis Distance, and Euclidean Distance). Distance measurement shall be applied on number with primary school facilities in Indonesia. The Davies Bouldin Index (DBI) is different from the cluster number test (k = 2-10) for each Distance Measure. The average DBI value in the type of measure (mixed measure) and numerical measurement (Mixed Euclidean Distance) is 0.54. The average DBI value in the type of measure (Bregman divergences) and numeric measurements (generalized IDivergence) is 0.66. The average DBI value is 0.77 for the measurement type (Bregman divergences) and numerical measurement (Squared Euclidean Distance). From the results, the measurement of distance with mixed measurement and the mixed Euclidean distance with the cluster number (k = 2), namely 0.269, have the best DBI value.

• Details
• Contents