PREDICTION AND ANALYSIS OF SOIL MACRONUTRIENTS USING MACHINE LEARNING TECHNIQUES

MACHINE LEARNING

Everyone has to be healthy and have access to enough crops and food as the world's population rises. The nation's economy benefits from crop output. Accurately estimating the percentage of macronutrients in the soil will assist in selecting the best crop to cultivate. Recent years have seen numerous advancements in everything from harvesting to product selection. The yield is increased when the correct crop is chosen for growing. Understanding the requirements for macronutrients is crucial for achieving optimal yield. The amount of NPK levels needed for various crops varies. The variables that must be taken into account are pH, temperature, humidity, and rainfall. Different machine learning models are often trained, tested, and validated. Prediction techniques include Decision Trees, AdaBoost Classifiers, XGB Classifiers, Random Forests, Logistic Regressions, SVM (Support Vector Machine) Classifiers, and KNN Algorithms. We can select the most effective model by contrasting the accuracy of several models. The creation of a user-friendly website is the suggested process. Inputs from the user include NPK levels, pH, temperature, humidity, and rainfall. Following the machine learning model's study, the website will take the macronutrients values as the input and suggest us the appropriate crop that may be cultivated in a given area as an output. We saw improved accuracy in the KNN Algorithm, SVM Classifier, and logistic regression during training, testing, and validation. The KNN algorithm's accuracy was determined to be 0.9886. We have used Google colab for training and testing the machine with datasets. We have created a user-friendly website by merging these two procedures.

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