Abstract

The agricultural sector plays an essential part in the financial progress of any country. The manufacture of agriculture is significantly compressed by rice, a crop developed all across the globe. Prompt prediction of diseases is essential for limiting their propagation and decreasing crop damage. However, physically analyzing crop illnesses in areas with massive agricultural regions and limited professionals is hugely complex. Using deep learning (DL) and machine learning (ML) models for analyzing illnesses of farm yields seems effective and appropriate for extensive applications. Federated learning (FL) has become a developing technology for data analysis for vast IoT applications. The manuscript proposes an IoT-Enabled Federated Learning for Crop Yield Prediction Using Machine Learning and Optimization Algorithms (IoTFL-CYPMLOA) technique in agriculture. The IoTFL-CYPMLOA model mainly focuses on enhancing the crop yield prediction model that improves agricultural production. Initially, the data pre-processing stage involves various steps such as categorical to numerical, handling null values, and normalization to clean, transform, and organize raw data into a suitable format. Furthermore, the XGBoost method is utilized for prediction. To improve the XGBoost model’s prediction performance, the parameter tuning process is performed by implementing the coot optimization algorithm (COA)method. The analysis of the IoTFL-CYPMLOA technique is examined under the crop yield prediction dataset. The experimental validation of the IoTFL-CYPMLOA technique portrayed results. Compared to recent approaches.

Authors

J. Jagadeesan, R. Nagarajan
Annamalai University, India

Keywords

IoT, FL, Coot Optimization Algorithm