SmartHarvest: Advancing Precision Agriculture Techniques for Enhanced Sustainable Crop Yield Management

Agriculture is a crucial pillar of India’s economy, providing livelihoods and engagement for a significant portion of its population. However, contemporary farmers encounter numerous difficulties, including enhancing crop yields, managing resources sustainably, and navigating an increasingly uncertain climate. This research examines machine learning (ML) practices that aid crop production planning by studying complex environmental and soil parameters like nutrients, temperature, humidity, and precipitation. Various ML models were examined, such as Logistic Regression (LR), RF (Random Forest), Gradient Boosting, SVM, NB (Naive Bayes), Decision Trees, and KNN (K-Nearest Neighbors), using a dataset containing 2,200 entries from different crops. Naive Bayes was the most precise model, reaching an accuracy of 99.55%, with Random Forest trailing closely behind at 99.32%. Gradient Boosting, Decision Trees, and KNN had high performance, achieving accuracies of 98.18%, 98.86%, and 97.05%, respectively. SVM and Logistic Regression achieved slightly reduced accuracies of 96.14% and 94.55%, respectively. When models are evaluated by precision, recall, F1-score, and accuracy, Naive Bayes consistently performed well in all areas. These results show how machine learning can also give agriculture an extreme facelift, enabling better crop management decisions. Transport (and spread) the technology and improve resource efficiency, cut waste & food security. The research emphasizes the importance of agricultural information in tackling modern farming obstacles.