Given the increasing pressures on global food production caused by a rapidly rising population and climate change, precise crop yield prediction has emerged as an important component of sustainable agriculture. This research delves into identifying and evaluating multiple crop yield prediction methodologies, focusing mainly on application with the Artificial Neural Networks (ANNs) model. The study highlights the problems with traditional statistical approaches like Linear Regression, and compare with them the capabilities of more modern machine learning methods especially ANN, highlighting their additional explanatory power and predictive accuracy. This research uses a dataset of 1,000,000 records with many features like region, soil type, crop type, and climatic variables to establish comprehensive basis for analysis. This study relies on mutual information gain to identify influential features. Our research takes advantage of rich dataset and adaptability of ANN model to reveal insights in optimizing agricultural output while enhancing food security. The proposed ANN model demonstrated superior performance with an R \(^{2}\) score of 0.9 and a Mean Squared Error (MSE) of 0.25, indicating high accuracy and minimal prediction error. In contrast, the Lasso Regression model achieved an R \(^{2}\) of 0.76 with an MSE of 0.39, while the Decision Tree Regressor performed better with an R \(^{2}\) of 0.82 but still had a higher MSE of 0.51, showcasing the ANN model’s overall efficiency. The work brings out this potential change that can be adopted by ANN in the prediction of agricultural yield to create better future prospects for precision farming. This work emphasises on ANN-based approaches that can be integrated into precision farming systems for predicting crop yield, offering scalable and efficient solutions for enhancing productivity.