Monitoring Soil Health Using Machine Learning Techniques for Improved Crop Yield Prediction

Prediction of soil nutrients is one of the important primary input management systems to enhance crop yield. Soil is a natural resource that can be found everywhere on Earth's surface. It is composed of minerals, organic matter, air, water, and other elements. Soil nutrients are chemical elements that plants need for growth and survival. There are various nutrients and factors on which soil health directly or indirectly depends, like nitrogen, phosphorus, potassium, pH, microbial activity, soil structure, moisture retention, iron, zinc, temperature, soil salinity, soil compaction, organic matter, etc. Also, farmers tend to have incomplete knowledge and information about proper fertiliser usage and their impact on the soil. As a result, sometimes they apply fertilisers in greater quantity than necessary, leading to an imbalance in soil nutrients. Hence, the work aims to monitor the health of soil based on various factors using machine learning techniques. In this work, 13 macronutrients that are needed for healthy plant growth and healthy soil are focused on. As a classification algorithm, light gradient boosting machine is used for the health prediction of soil. The proposed model gives an accuracy of 95.56%. Such a high level of accuracy demonstrates the model's ability to identify complex relationships between soil nutrients and health indicators. Furthermore, the insights gained from the proposed work can empower farmers to make smart and sustainable soil health management decisions. By accurately assessing soil health, farmers can optimise fertiliser use, enhance crop yield, and promote sustainable agricultural practices. The given approach not only contributes to increased productivity but also helps in preserving soil integrity for future generations.