Application of Indonesian's National Standard for Grouping of Bali Cattle with Cluster and Biplot Analysis

Research on the application of the Indonesian National Standard (SNI) of Bali cattle by cluster analysis and biplot aims to provide a visual picture in the form of tables and graphs, so that it is easier and faster and more communicative in making decisions, whether the cows studied are included in class I, class II, or class III based on SNI Bali cattle. This study was conducted on 70-year-old adult female cows of 70 animals raised at the Integrated Farming System (Simantri) in Badung regency. The data obtained were analyzed by cluster analysis and biplot, as variables were shoulder height, body length and chest circumference, while as objects were 70 adult cows and 3 classes of Balinese cattle based on SNI of Bali cattle as object identifiers. The results obtained that the application of SNI for Bali cattle can be done by cluster analysis and biplot and both analyzes give the same results to the grouping of Bali cattle objects based on SNI for Bali cattle. Grouping by cluster analysis is easier to see based on the cluster membership obtained, whereas with biplot analysis provides additional information about correlations and diversity between variables.


I. INTRODUCTION
as minimum requirements for Bali cattle breeding [1].
Broadly speaking there are two ways of presenting data that are often used are tables or lists and graphs or diagrams [2].
Demonstration of data with pictures (graphs) has been long and often done, and is highly recommended in an analysis. Usually the display of data in this form is more favored than the presentation in the form of tabulation of numerical data or in the form of narration, because it is more interesting, more informative because it can provide more information and is more communicative so that it is easier to understand and can be said tobe more artistic in form, composition , and colors so that it is more beautiful [3].
Multivariate analysis is a statistical analysis relating to the description or interpretation of data which involves many variables and objects together [4]. One such interpretation is the grouping of data [5].
Data grouping is organizing a large data set by dividing the data into several groups (groups). The groups formed will be able to explain the similarities and differences of the entire data studied [6]. The purpose of grouping data is to facilitate the process of analysis and interpretation of big data by dividing the data into several groups. There are several techniques in grouping a data, data can be grouped based on variables and can also be based on the object of the data.
One analysis that is often used for this grouping is cluster analysis.
Cluster analysis is a multivariate analysis that aims to group objects from the data studied based on the similarity of the characteristics they have [6]. The similarity of these characteristics is usually measured using a measure of proximity between objects which can be a measure of similarity or dissimilarity.
Cluster analysis is a grouping of objects or cases into smaller groups where each group contains objects that are similar to each other [7]. Cluster analysis (cluster analysis) is one of the statistical methods that can be used to carry out a grouping process. In its grouping we use a measure that can explain the closeness between data to explain the simple group structure of complex data, which is a measure of distance. A measure of distance that is often used is the size of the Euclidean distance [8].  Biplot is an exploratory method of analyzing variable data that can provide a graphical picture of the closeness between objects, diversity of variables, correlations between variables, and the relationship between variables and objects. In addition, biplot analysis is used to describe the relationship between variables and objects that are in high-dimensional space into lowdimensional space that is two dimensions [13]. Graphical simulations are expected to get a picture of objects, for example the closeness between objects, and descriptions of variables, both in diversity and correlation, as well as the relationship between objects and variables [12].
Both multivariate analyzes have their respective advantages, cluster analysis can be used in grouping data in large numbers of objects as well as with different scale measurement variables ranging from nominal to intervals [6]. Biplot analysis is able to directly display the predominant variables or variables that are the most dominant of a group of objects formed on the results of the biplot analysis display [14].
grouping of biplot analysis produces a better percentage of accuracy than clustering cluster analysis. But in general, it cannot be said that biplot analysis is better than cluster analysis in grouping data and vice versa [15].