CLASSIFICATION OF BEANS USING PRINCIPAL COMPONENT ANALYSIS

ABSTRACT Classification of beans is a crucial task in agriculture, food science, and the food processing industry. Beans, being a significant source of protein, fiber, and various nutrients, play a vital role in global food security and nutrition. The manual classification of beans based on their physical and nutritional properties can be prone to errors, time-consuming, and costly. With the growing demand for high-quality beans and the diverse types available, there is a need for an accurate, efficient, and automated system to classify beans. The challenge is to develop a reliable method that can handle the complexity of beans features and deliver consistent classification results. Hence, Principal Component Analysis (PCA) provides a promising solution. Therefore this study focuses on the classification of beans using PCA in the context of Nigerian agriculture. The research aims to develop an accurate and efficient method for categorizing different bean varieties based on their physical characteristics. The study employed PCA to reduce the dimensionality of a dataset containing various bean features. K-means clustering was then applied to the PCA-transformed data to group beans into distinct categories. The analysis was performed using R programming language, utilizing libraries such as prcomp, ggplot2, and caret for data manipulation, visualization, and model implementation. PCA effectively identified key attributes such as size, shape, and texture for distinguishing between bean types. K-means clustering on the PCA-reduced data revealed seven distinct clusters. The model showed high precision for some bean types such as BOMBAY and SIRA but lower accuracy for others which includes DERMASON and SEKER.