Countries population determination to test rice crisis indicator at national level using k-means cluster analysis

The Indonesian Operations Research Association (IORA) - International Conference on Operations Research 2016

Hidayat, Y.^1 ; Purwandari, T.^1 ; Sukono^2 ; Ariska, Y.D.^3

Department of Statistics, Universitas Padjadjaran, Indonesia^1

Department of Mathematics, Universitas Padjadjaran, Indonesia^2

Graduate School of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Padjadjaran, Indonesia^3

关键词: Data population;    Differentiators;    Hierarchical cluster analysis;    K-means cluster analysis;    Multi variate analysis;    Number of clusters;    Purchasing power;    Reference material;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/166/1/012023/pdf
DOI  :  10.1088/1757-899X/166/1/012023

来源: IOP

【 摘 要 】

This study aimed to obtain information on the population of the countries which is have similarities with Indonesia based on three characteristics, that is the democratic atmosphere, rice consumption and purchasing power of rice. It is useful as a reference material for research which tested the strength and predictability of the rice crisis indicators Unprecedented Restlessness (UR). The similarities countries with Indonesia were conducted using multivariate analysis that is non-hierarchical cluster analysis k-Means with 38 countries as the data population. This analysis is done repeatedly until the obtainment number of clusters which is capable to show the differentiator power of the three characteristics and describe the high similarity within clusters. Based on the results, it turns out with 6 clusters can describe the differentiator power of characteristics of formed clusters. However, to answer the purpose of the study, only one cluster which will be taken accordance with the criteria of success for the population of countries that have similarities with Indonesia that cluster contain Indonesia therein, there are countries which is sustain crisis and non-crisis of rice in 2008, and cluster which is have the largest member among them. This criterion is met by cluster 2, which consists of 22 countries, namely Indonesia, Brazil, Costa Rica, Djibouti, Dominican Republic, Ecuador, Fiji, Guinea-Bissau, Haiti, India, Jamaica, Japan, Korea South, Madagascar, Malaysia, Mali, Nicaragua, Panama, Peru, Senegal, Sierra Leone and Suriname.