Robust Fast Minimum Covariance Determinant Elastic Net HJ Biplot Analysis for Mapping Cabbage Yields in Malang

Authors

  • Alifiandi Rafi Muhammad, Ni Wayan Surya Wardhani, Atiek Iriany, Prayudi Lestantyo

DOI:

https://doi.org/10.17762/msea.v71i4.613

Abstract

The crop yields mapping is necessary to map certain commodities so that they can be followed up in increasing the production of these commodities. A multivariate analysis to determine the characteristics of production data is biplot analysis. In previous studies, the biplot analysis has applied an elastic net approach to reduce data dimensions. However, it has not been able to handle the outliers. An example of an estimator which is resistant to outliers is Fast-MCD. This study aims to compare the value of the goodness of fit of the biplot using HJ Biplot, Elastic Net HJ Biplot, and the combined Robust Fast-MCD Elastic Net HJ Biplot method. The time point (year) is used as a variable to consider the relationship between cabbage yields each year. The data used is cabbage production in Malang area from 2017 – 2021. The data analysis process uses the RStudio application with three special packages: rrcov, SparseBiplots, and sparsepca. The results of this study indicate that the combined method can increase the goodness of the biplot by 0.15% compared to the HJ Biplot and 0.06% compared to the Elastic Net HJ Biplot. This biplot also shows that the highest production is obtained in shifts in several regency which is indicated by the changing year vectors in a clockwise direction.

Downloads

Published

2022-08-29

How to Cite

Alifiandi Rafi Muhammad, Ni Wayan Surya Wardhani, Atiek Iriany, Prayudi Lestantyo. (2022). Robust Fast Minimum Covariance Determinant Elastic Net HJ Biplot Analysis for Mapping Cabbage Yields in Malang. Mathematical Statistician and Engineering Applications, 71(4), 1159–1167. https://doi.org/10.17762/msea.v71i4.613

Issue

Section

Articles