Abstract: Agriculture is the hub of international food security and economic stability. Farmers are, however, faced with unpredictable climatic patterns, soil quality variations, and inadequate access to accurate predictions of yield. The traditional forecasting techniques rely on manual data and complex interpretations and, hence, are less accessible to farmers. This paper suggests an Agricultural Yield Forecasting System with machine learning concepts to provide accurate predictions of yield with fewer input parameters. The system employs Lasso Regression, Elastic Net (ENet), and Kernel Ridge Regression, with stacking procedures to achieve maximum accuracy. Our method demonstrates improved efficiency, accuracy, and accessibility and provides a user-friendly web-based solution, which can be further developed as a mobile app with regional languages.
Keywords: Agricultural Forecasting, Machine Learning, Crop Yield Prediction, Regression Models, Data-Driven Farming
Copyright Form