MACHINE LEARNING BASED PREDICTIVE MODELLING TO ANALYSE THE MOISTURE LEVEL OF SOIL FOLLOWED FOR EFFECTIVE DROUGHT CONTROL

Moisture level of soil is one of the main factors in agriculture productivity and other organic applications. Moisture level in soil is not constant always and it varies with rainfall, temperature and other environmental factors. It is essential to measure the moisture level of soil to avoid drought condition of soil. Monitoring and measuring the moisture level in soil is tedious and error prone task when done manually and it is much time consuming but still results with less accuracy. In order to address this issue and improve the agriculture productivity and other ecological purpose machine learning based predictive modelling is proposed here for effective water resource management and drought control. Supervised learning kind of machine learning is adopted here which involves training and testing to predict the moisture level of soil. The training database is loaded with the values measured from the moisture sensor at regular intervals and sent via cloud. The data base must be loaded with necessary details such as date, moisture level, type of soil and the data from the database is used as test data to get the predicted response based on previously trained data