Radial Kernel Truncated Gradient Margin Boost Classification for Efficient Crop Yield Prediction

Agriculture involves cultivating land, growing crops, and raising animals for food, fiber, and other products essential for human life. In crop yield prediction, agriculture involves forecasting the amount of crop production from a given area of land. This process utilizes various methods, including historical data analysis, weather forecasting, soil conditions, and crop management practices. Accurate yield predictions help farmers make informed decisions about resource allocation, optimize crop management, and manage risks related to climate and market fluctuations. Several machine learning techniques have been developed, but timely yield prediction remains a challenging issue. A novel method called Radial Kernel Truncated Gradient Margin Boost Classification (RKTGMBC) has been developed for accurate crop yield prediction, achieving higher accuracy and lower time complexity. The main aim of the RKTGMBC method is to perform several processes such as data acquisition, preprocessing, and feature selection. Following this, crop yield prediction is performed using the selected features through an ensemble classification method. In the RKTGMBC method, the number of selected relevant features is used as input for the Truncated Gradient Margin Boost ensemble classification method. This method employs the radial basis kernel perceptron as a weak learner to analyze the data samples and provide final classification results. The Margin Boost ensemble classification method combines the results of the weak learners and applies the Truncated Gradient method to provide stable output classification results by minimizing or maximizing the margin to reduce error. In this way, accurate crop yield prediction is achieved with minimal computational time. Experimental evaluation considers factors such as crop yield prediction accuracy, precision, recall, F1 score, and prediction time with respect to the number of data samples. The quantitatively analyzed results indicate that the proposed RKTGMBC method achieves higher crop yield prediction accuracy with minimal computation time compared to conventional techniques.