Ensemble Machine Learning Algorithms for Crop Types Classification Based on Intelligent Detection of Soil Properties
Abstract
In this study, a model for analyzing soil fertility in agriculture using machine learning technologies and recommending suitable crops was developed. A dataset based on various soil properties was analyzed using modern machine learning algorithms, in particular, KNN and Gaussian Naïve Bayes. The research compiled a dataset containing key factors affecting soil fertility and used these factors to classify crop types. To improve the performance of the algorithms, the study combined KNN and Gaussian Naïve Bayes through an ensemble learning technique known as the voting method. The findings showed that the soft voting approach delivered greater accuracy than the individual models. This method proved to be highly effective in providing dependable recommendations for crop planting based on soil characteristics.