A Machine Learning-based Agricultural Yield Forecasting System for Predicting Crop/Plant Yield before Planting the Crops/Plants
R. Arun, T. Saraswathi, D. Meeha, G. Sugeerthi
Abstract
One of the main causes of the current decline in agricultural work in India is farmers’ agricultural losses. Due to this, predicting crop yield before plantation is becoming one of the overriding strides of this era. Even before planting a crop, it is very important to predict the yield due to the frequent changes in the parameters involved. Usually, crop/plant production will be based on various geographical strategies, while analyzing such parameters using machine/deep learning algorithms helps to choose the crops that correspond to the geographical region. It will serve as a preventative measure that helps to reduce agricultural losses while also enhancing the productivity and quality of the crops that are produced. In this regard, various Machine learning models and Deep Neural Network models are employed to analyze the various geographical parameters that influence agricultural production. In this study, yield prediction is calculated in terms of state, district, area, and season for every crop/plant wise. It helps the farmers to select appropriate crops at the right place and time. The preciseness of the prediction is evaluated using various parameters such as accuracy, mean absolute error, and mean square error. Among the utilized machine/deep learning models, the Deep Neural Network model achieved Mean Absolute Error, and Mean Squared Error of 2.16% & 5.14%, and 3.02% & 8.56% for Andhra Pradesh and Punjab state dataset, respectively.