Improving and Prediction of Efficient Soil Fertility by Classification and Regression Techniques

Authors

  • Yogesh R. Shahare , Vinay Gautam

Abstract

Soil is a significant parameter for the production of crops prediction. Soil analysis can benefit farmers and land Analysts are allowing higher crop yields by Arrangements. . This report has been prepared with the intent of predicting soil fertility by applying a decision tree classification, regression and random forest classification and regression techniques. Soil data were collected from District Soil samples from various Maharashtra districts for the experimental establishment. In order to evaluate the efficiency of each studied technique, few parameters have been assessed for performing the random forest and decision Classification such as OOB error rate, AUC accuracy, Error rate, recall, precision, f-score and regression techniques such as, mean squared error, r2 score, and error rate. The most well-known studies have been conducted precise soil prediction techniques using machine learning approach of training, testing and regression, classification methods. The experimental results predicted that decision-making tree and random forest are the most reliable methods of soil fertility prediction

Keywords- Decision tree classification and Regression, Random Forest classification and Regression,  Micro soil nutrients, Macro soil nutrients

Published

2020-12-31

Issue

Section

Articles