ERFC: Crop Prediction-Based Agricultural Environment Using Enhanced Random Forest Classification

Agriculture is absolutely essential to human existence. The prediction of crops is especially important in agriculture and mostly dependent on the soil and environmental factors, such as temperature, humidity, and rainfall. Prior research has achieved accurate classification through appropriate feature selection; however, the prediction of feature selection is time-consuming. In this paper, the ERFC algorithm—a novel Enhanced Random Forest Classification (ERFC)—is presented. By combining the Improved Recursive Feature Selection (IRFS) algorithm with classification, a novel method known as the ERFC algorithm is produced. The present feature selection methods greatly enhance the probability that redundant features will appear in the final subset; yet, finding and removing them can improve the classification accuracy considerably in most cases. IRFS and the ERFC methodology are two approaches used to address this issue. Comparing the proposed ERFC algorithm to other classification algorithms like Naive Bayes, Decision Tree, Support Vector Machine (SVM), and Random Forest algorithms reveals that it has a high precision, recall, accuracy, and F1-score ratio.