Analysis of eigenvalues convergences using principal componen analysis, Case study : Data on land cover change in West Java

• Main Authors: Sari, Kunia Novita, Pasaribu, Udjianna Sekteria, Siahaan, Darman, Ilmi, Nisa Fadlilah Fathul, Viridi, Sparisoma
• Format: Conference Proceeding
• Language: English
• Subjects: Complex variables; Data analysis; Earth surface; Economic factors; Eigenvalues; Independent variables; Land cover; Principal components analysis; Random variables
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0175727

Land cover is the condition of the observed biophysical appearance of the earth’s surface. From 2005 to 2010, land cover in West Java has undergone various changes in land cover caused by physical and economic factors. The causal factor of the large number of land cover changes resulted in very large and complex computations. Principal component analysis was carried out to simplify this condition. The purpose of this analysis is to reduce the dimensions of correlated random variable into uncorrelated or independent. The variables used in this study are the slope of the land, curvature, height of land, distance to the capital, distance to Bekasi City, and distance to Bandung City. After the analysis, the six variables can be represented by three principal components with an initial data absorption of 78%−83% based on each number of observations. Conducted a study with a different number of observations, namely 100, 200, 300, 400, 500, and 4242837 pixels, to see the convergence of the eigenvalues for each number of obervations with the population.