Soil Moisture Prediction Using Machine Learning Techniques

Agriculture is very important component of India's economy. In the atmosphere, soil moisture is a major factor and acts an imperative in disaster predictions, environmental monitoring and water shortage. In this work, support vector regression and decision tree techniques have been applied on the given datasets. The data used in this experiment is classified into two different parts: soil moisture data and meteorological data. The data is collected from different stations for the prediction of soil moisture in advance. The performance of decision tree technique is assessed by using mean square error and coefficient of determination. From the experimental results, it has been concluded that decision tree provides better results than the support vector regression. This provides idea to the researchers in the prediction of soil moisture content.