PRINCIPAL COMPONENT ANALYSIS ON SOIL CHEMICAL DATA [Case Study: Palm Oil Garden in West Sulawesi]

• Main Author: RIFQI NADZORI (NIM: 10114088), AINUR
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/25231
• Institution: Institut Teknologi Bandung

Palm oil is one of the non-oil export commodities that is important and contributes to the Indonesian economy. The productivity of oil palm plantations in Indonesia today must be improved and maintained with a good management such as maintenance activities. One of the activities in maintenance that requires management is fertilization activities. Fertilization aims to increase soil fertility so that with good soil fertility the productivity of oil palm will increase. In this final project will be carried out the principal component analysis on soil chemical data. Principal component analysis (PCA) is a statistical method to reduce data dimensions that have been widely used both in terms of data processing, machine learning, and image processing or signal processing. This method is used to determine what factors are most influential on soil fertility of oil palm. Principal component analysis is carried out by singular value decomposition (SVD). Data comes from an oil palm plantation in West Sulawesi at depths of 20 cm to 40 cm which contains soil nutrients. Based on the results of principal component analysis, it was found that cation exchange capacity (CEC), C-Organic availability, nitrogen (N), and K2O compounds were the most influential factors in soil fertility in the garden. These results are expected to help manage soil fertility more effectively and efficiently such as determining the right type of fertilizer to spur the growth of oil palm and increase productivity.