HarvestVision: A Machine Learning-based Platform for Accurate Crop Yield Prediction

HarvestVision is a machine learning tool used to visualize precision crop yield data and provide intelligence to those in the agricultural industry. With the help of an XGBoost algorithm, the platform predicts crop yields relying on historical yield, climate, and region thus offering useful insights in the farmers' decision-making. Comparison of results between XGBoost and other machine and deep learning algorithms like Random Forest, Lasso, CNN, LSTM in terms of precision, and scalability have been done. The presented results show that XGBoost delivers the highest degree of accuracy while maintaining reasonable computational time, thus, allowing to predict with precision across different crops and regions. The platform comprises of areas such as the prediction history module for analyzing the trends as well as the chatbot module for the general questions about the crop yield and farming. Through HarvestVision the concept of applying advanced machine learning with simple user-friendly resources is demonstrated, suggesting how elevation of agri-management can enhance productivity for buyer-sellers and producers of various scales.