Soil Moisture Prediction Using Machine Learning

Abstract

Prediction of soil moisture in advance is useful to the farmers in the field of agriculture. In this paper we have used machine learning techniques such as multiple linear regression, support vector regression and recurrent neural networks for prediction of soil moisture for 1 day, 2 days and 7 days ahead. These techniques were applied on three different datasets collected from different online repositories. The performance of the predictor is evaluated on the basis of mean squared error(MSE) and coefficient of determination (R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> ). The comparison result shows that multiple linear regression is superior providing MSE and R2 of 0.14 and 0.975 for 1 day ahead, 0.353 and 0.939 for 2 days ahead, 1.59 and 0.786 for 7 days ahead.