A review of soil analysis and classification to improve crops in agriculture using machine learning algorithm

Abstract

One of the most crucial aspects of our society is agriculture. For agriculture to be successful, soil is essential. Each type of soil has a unique composition. These chemical properties of the soil have an impact on crop growth. It is important to choose the right crop for that specific type of soil. Machine learning algorithms are able to classify the data from the soil series. to predict which crops are best suited to the soil type and climate of a certain area, the findings of this categorization can also be paired with crop datasets. Both the crop and the soil datasets are used. The files include geographical and chemical characteristics of soil and crop features. The classification of soil series data and the prediction of appropriate crops can be done using algorithms like SVM and the assembling approach. For this study, 15 villages in the near area of Kampala, Uganda, provided 1200 soil samples with chemical, physical, and biological characteristics. Overall, the Random Forest model produced results with 99% accuracy for cellulose enzyme activity and 94% accuracy for N-acetyl-glucosaminidase enzyme activity. Based on evaluation performance parameters and the best performance accuracy algorithms, recommendations for crops have been made.