Crop-RecFIS: Machine Learning Classifiers for Crop Recommendation and Feature Importance Scores Calculation

• Published in: 2023 14th International Conference on Computing Communication and Networking Technologies (ICCCNT) pp. 1 - 6
• Main Authors: Ferdib-Al-Islam, Sanim, Mostofa Shariar, Mehedi Hasan, Khan, Alam, Md Magfur, Walid, Md. Abul Ala, Islam, Md. Rahatul
• Format: Conference Proceeding
• Language: English
• Published: IEEE 06.07.2023
• Subjects: Correlation; Crop Recommendation; Crops; Decision Tree; Decision trees; Estimation; Feature Importance Score; Machine learning; Machine learning algorithms; Seeds (agriculture); Visualization; XGBoost
• ISSN: 2473-7674
• DOI: 10.1109/ICCCNT56998.2023.10307833

Crop recommendation estimation is a critical topic in agriculture. In the past, crop recommendation was made by considering a farmer's familiarity with a particular field and seed. The farmers' failure to choose the appropriate crop for cultivation is a significant and dangerous setback in crop productivity. The outcomes of several machine learning models suggest the crop is based on the input factors with high performance. This study aims to propose crop recommendation models based on machine learning approaches. The primary goal of this research was to assist farmers in selecting the best-suited crop for their area. In this study, 22 crops have been classified using several machine learning algorithms. The feature importance score has also been calculated to identify the feature's effect on the machine learning models. Among the machine learning models, both the decision tree (DT) and XGBoost models achieved the highest 99% accuracy, precision, and recall, outperforming existing works.