Soil Moisture Prediction Using Machine Learning Techniques

Abstract

Although - Soil moisture is the main factor in agricultural production and hydrological cycles, and its prediction is essential for rational use and management of water resources. However, soil moisture involves complicated structural characters and meteorological factors, and is difficult to establish an ideal mathematical model for soil moisture prediction. Prediction of soil moisture in advance will be useful to the farmers in the field of agriculture. In this paper, we have used machine learning techniques such as linear regression, support vector machine regression, PCA, and Naïve Bayes for prediction of soil moisture for a span of 12 to 13 weeks ahead. These techniques have been applied on four different datasets collected from 13 different districts of West Bengal, and four different crops (Potato, Mustard, Paddy, Cauliflower) collected over the span of about 1st January 2020 – 30th March 2020. The performance of the predictor is to be evaluated on the basis of F1-Score.