Comparative Analysis of Different Machine Learning Prediction Models for Seasonal Rainfall and Crop Production in Cultivation

Abstract

Agriculture is one of the strengths of India, from the last few years, gradually the agriculture growth is going downwards in other side the population growth is upwards. Reason for agricultural downward growth depends on so many parameters. The rainfall is one of the main parameters which affects the crop yield. Because of this, the farmers are also facing the loss. If they know this information in prior, the farmers can plan accordingly the type of crop suited for the particular season and it helps the farmer to get good profit out of it. Machine learning scientific and statistical methods are used for predicting the rain fall and crop yield. Kharif and Rabi are two seasons taken for analysis. The regressor predicting models are constructed to predict the seasonal rainfall and crop yield. This study primarily focuses on seasonal crop production prediction, which is dependent on rainfall. The different types of machine learning regression method are used to achieve better results. The performance of comparison models is evaluated using different metrics. Finally, the linear regression and Bayesian linear regression models comparatively produce the best result in terms of accuracy for rainfall prediction. The boosted decision tree regression model is achieving the better result for crop prediction.