SELECTION OF CROPS BY REGION USING MACHINE LEARNING CLASSIFICATION ALGORITHMS

Abstract

The article elucidates research into a prediction approach which is based on machine learning algorithms for selecting types of crops that can be planted for forthcoming seasons, in accordance with climatic conditions in agriculture of a certain region. The research investigated the main mathematical core, basic concept and hyper-parameters of the algorithms designed for creating an AI model, analyzed individual models using the dataset selected for the research, and retrieved required findings. The findings were compared with results of other algorithms based on a number of indicators including F1-score, Recall, Accuracy. While studying the classification algorithms, they were considered in terms of efficacy – as to what kind of problems the algorithms would be most effective for; and the article provided specific comments on this issue. The software was developed in the Python programming language in view to ensure clear visualization of the comparative analysis of the research findings. As well as, the results section contains graphical figure of each input/output attributes’ relationship.