A Network-Centric Web Application for Dynamic Crop and Fertilizer Decision Support

Abstract

The economy depends heavily on agriculture, and improving productivity and decision-making requires precise crop forecasting. In order to suggest the best crop, this project offers a crop prediction system that takes into account important environmental parameters like phosphorus, nitrogen, potassium, rainfall, humidity, temperature, and pH level. To examine the relationship between soil characteristics and environmental factors, the system makes use of machine learning algorithms, including Random Forest, XGBoost, and K-Nearest Neighbors (KNN), which are known for their high accuracy and resilience. These models help farmers make well-informed decisions by processing large datasets and producing accurate crop recommendations. By minimizing overfitting and effectively managing non-linear data, these models guarantee improved performance. Enhancing agricultural productivity, maximizing resource use, and promoting sustainable farming methods are the goals of this system.