Published Paper Details:

Agriculture is a backbone of the global economy, and empowering farmers with advanced technology and tools can enhance productivity and sustainability. Traditional agricultural methods were based on manual tools, natural fertilizers, practices like crop rotation and rainwater harvesting to sustain productivity and soil health. Farmers used their predictive, observational knowledge for weather, crop prediction and recommendation natural pest control remedies, though these methods lacked scalability and resilience against environmental changes. Machine learning offers a promising solution by providing the prediction and recommendation capabilities by analyzing vast datasets, including soil, weather, and crop patterns, to forecast yields, recommend and predict suitable crops, and optimize fertilizer usage. This paper proposes a machine learning-based crop, fertilizer recommendation and prediction, weather and rainfall prediction. The system uses 4 features for crop prediction, 8 features for crop recommendation, 9 features for fertilizer recommendation and 7 features for yield prediction. Along with tools for farmer's assistance like news feed, chatbot support. We evaluate four models--Decision Tree Classifier, Random Forest Classifier, Random Forest Regressor and Statistical Analysis --showing that the models achieve high accuracy by leveraging data-driven patterns, feature importance analysis to optimize predictions and recommendations for agricultural needs.