Comprehensive Analysis of Artificial Intelligence based Crop Recommendation and Soil Analysis

In recent years, crop recommendation has become a significant research area that utilizes environmental characteristics of humidity, temperature, soil Potential of Hydrogen (pH) and rainfall to determine the appropriate crops for production. The prediction of crop yields is a difficult task that depends on weather, soil conditions, crop-specific variations, and environmental factors. The automated crop recommendation approach is essential in helping farmers make prior knowledgeable decisions regarding production and crop cultivation. In this survey, machine learning (ML) and deep learning (DL) algorithms are analyzed for soil analysis and crop recommendations. The performance of the existing methods is estimated based on various performance measures such as accuracy, recall, f1-score, precision, sensitivity, kappa, area under curve (AUC), (RMSE), mean absolute percentage error (MAPE), mean square error (MSE) and mean percentage error (MPE). This survey concludes that various soil analysis and crop recommendation approaches overcome drawbacks such as high time consumption and inaccurate crop outcomes.