A Framework for Rainfall Forecasting based Crop Analysis using Deep Learning

More than half of India's arable land is susceptible to drought, making the country's agriculture perpetually insecure. Crop productivity and agricultural investment are both negatively impacted by the monsoon's poor performance. Literature states that the existing statistical and numerical procedures often fail to meet the present requirements and strategies for recognizing the important input data for designing efficient rainfall and crop yield forecasting models. In this research, Adaptive Moment Optimization (AMO) based Long Short-term Memory Neural Networks (LSTM-NN) are developed for accurate seasonal rainfall forecasting. The AMO is used to find the relevant features for rainfall prediction and LSTM is used to predict the rainfall. The AMO has a higher learning rate and selects the relevant features that help to progress the performance of the model. The investigational outcomes demonstrated that the proposed strategy accomplished nearly 95% of accuracy in all cases.