Crop Yield Prediction Using Machine-Learning
Crop Yield Prediction Using Machine-Learning
DOI:
https://doi.org/10.56536/jicet.v4i2.136Abstract
This study examines the utilization of machine learning techniques for forecasting agricultural yields. This is critical because it helps to improve farm productivity and food security. Agriculture is the backbone of Pakistan’s economy, and accurate crop yield prediction can greatly assist in economic planning and allocation of resources. “Several ML algorithms, including K-Nearest Neighbor, Lasso regression, Ridge Regression, Decision Tree, and Linear Regression,” were used in this research to forecast crop yield using historical data and climate variables. Among all the models developed, the KNN algorithm was the most effective, with MAE=108023825.91 and R2=0.98494, meaning that it had the highest accuracy. These results suggest that ML has a large potential to give good predictions about crop yields, which would be useful for decision-making in agriculture and crop management or strategy development. The paper also enlightens on the limitations of some algorithms, like Decision Trees, known for their low-performance abilities; it suggests other ensemble models as better alternatives to increase predictive power. Machine learning integration with big data platforms and advanced computational methods offers a promising approach toward smart agriculture, ensuring food security.