AN IOT BASED INTELLIGENT DECISION SUPPORT SYSTEM FOR SMART FARMING USING MACHINE LEARNING TECHNIQUES
Farmers and agricultural executives rely on crop production and drought predictions to aid agriculture-affected regions all over the world. The prediction of Drought is crucial for early warning and reducing the effects of drought on agricultural yield. Drought forecasting research aims to enhance predictability skills and deepens the understanding of the physical mechanics of drought by combining all available predictability sources. In this paper, an intelligent decision support system for smart farming is proposed along with wrapper-based feature selection and support vector machine classifier is used for the productivity of crop and the prediction of the drought. The results are evaluated using three real-time datasets and compared with five existing prediction models namely K-Nearest Neighbour, Naïve Bayes, Back Propagation Neural Network and Wrapper based PART. The productivity of the crop's sorghum, jowar and sugarcane are used for the prediction and the Experiment outcome reveals that the proposed wrapper-based feature selection and wrapper-based classification is best suitable for drought prediction and productivity of crops.
Machine Learning, Agriculture, Drought Forecasting, IoT in Agriculture, Modernization of Agriculture, Intelligent Farming.