Crop Yield Prediction Data Analytics in Indian Agriculture Using Deep Learning

Abstract

India is a nation where agriculture and industries associated with it are the main sources of income for the populace. The country's economy primarily depends on agriculture. It is also one of the nations that experience severe natural disasters like floods or droughts, which ruin the crops. The current system uses regression approaches to estimate yield, such as Kernel Ridge, Lasso, and ENet algorithms, and it also employs the idea of stacking regression to improve the algorithms' performance. Utilise technology like data analytics and machine learning to analyse and mine this agricultural data to produce results that will be valuable to farmers for more productive and efficient crop yields. We suggest creating efficient methods to forecast agricultural yield under various climatic situations, which can assist farmers and other stakeholders in making knowledgeable decisions regarding agronomy and crop selection. The DNN algorithm, Multilayer Perceptrons (MLP), was employed. Additionally, the DL (Deep Learning) model's time and space complexity will increase with the addition of new characteristics that have minimal impact on the model's performance. The findings show that compared to the current classification technique, an ensemble technique provides more accurate prediction.