PREDICTION AND ANALYSIS OF SOIL MOISTURE USING MACHINE LEARNING TECHNIQUES

MACHINE LEARNING

As the world population grows, health and adequate crop and food production are important for everyone. Crop production positively affects the country's economy. Predicting the moisture content present in the soil will help in choosing better crop for cultivation. Many advances have been made in recent years, from product selection to harvest. Selecting the right crop for cultivation increases the yield. For maximum yield, the understanding of macronutrient requirement is important. Different crops require different amount of NPK values. Temperature, humidity, rainfall, and pH are the factors need to be considered. We tend to train, test, and validate various machine learning models. Decision Tree, AdaBoost Classifier, XGB Classifier, Random Forest, Logistic Regression, SVM (Support Vector Machine) Classifier, and KNN Algorithm can be used for predicting purposes. By comparing the accuracy of various models, we can choose the efficient model. The proposed methodology is to design a user-friendly website. The user can provide inputs such as NPK values, temperature, humidity, rainfall, and pH. After the analysis by the machine learning model, the website will display the moisture content and suitable crop that can be grown in a particular region. On training, testing and validation we got higher accuracies in Logistic regression, SVM Classifier and KNN Algorithm. The accuracy of KNN algorithm came out to be 0.9886. For moisture analysis we have trained and tested using Neural Network pattern Recognition in MATLAB. By combining these two processes, we have designed a user-friendly website.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.