Classification of Beans Using Principal Component Analysis: An Unsupervised Learning Classification Approach

Abstract: Manual classification of beans is prone to errors and inefficiencies, necessitating an automated system for accurate and consistent results. This study focuses on the classification of beans, an essential task in agriculture and food processing, using Principal Component Analysis (PCA) to handle the complexity of beans' features. The research aimed to develop an efficient method for categorizing Nigerian beans varieties based on physical characteristics. PCA was employed to reduce the dimensionality of a dataset containing features like size, shape, and texture, followed by K-means clustering for categorization. Using the R programming language and libraries like prcomp, ggplot2, and caret, the analysis identified seven distinct clusters. The model achieved high precision for some beans types (such as BOMBAY and SIRA) but lower accuracy for others (such as DERMASON and SEKER). Overall, the study demonstrated the effectiveness of PCA in simplifying datasets and enhancing classification accuracy.