Authors : D.A. Jakkan,Chandrasekhar Sakode,Pradnya H. Ghare
Page Nos : 73-77
The agricultural sector plays a major role in a country within the economy of the nation. In recent years, many innovations have also been introduced in the field of the agricultural industry. Serious research is happening all over the world, with the invention of IoT, Big Data Technology, and Machine learning how the farmers and even the Govt. Entities will help improve this sector with rising demand for agricultural products worldwide. Various soil properties like ph level, total nitrogen, phosphorus, potassium,carbon contents in soil impactson crop production in the agriculture sector. Methods for rapid and accurate soil tests are needed for detecting levels of above index properties of soil attributes which help the farmer to take correct decision for crop production. Mid-Infrared Spectroscopy (MIR) is one such technique used to identify the classes and properties of soils with high precision. In this research work,we are using machine learning model for analysis and prediction of Potassiumproperty by using Mid-infrared spectroscopy dataset collected from a various sample of California state in the United States. Our results show that the decision tree regression and random forest regression model gives better accuracy and through that,we can predict the Potassiumlevel of the soil for helping the farmer for getting right kind of crop in that soil sample.