Crop Suitability Prediction and Fertilizer Recommendation Using Classification Techniques

摘要

The backbone of the Indian economy is agriculture. In India, the majority of people depend on agriculture. The rise of the crop yield in agriculture is mainly based on selecting the appropriate crop suitable for the soil. Existing methods used rough set-based approaches for prediction of crop suitability. The rough set-based approaches did not embody the process of uncertainties present in the accumulated data. For this reason, fuzzy-based approach is used to handle the uncertainties during preprocessing the data in addition to rough set approach. The mixed approach of rough set and fuzzy is used to produce a classification model to aid farmers in taking decisions for crop cultivation. The fuzzy model is used for feature selection and the rough set-based approach is used in rule induction. The proposed method is compared with the rough set-based approach, Support vector regression and random forest regression are found to be more accurate.