Intelligent IoT-Based Data Analytics System for Precision Farming Using Regression Techniques

Abstract

Increasing global demands for sustainable agricultural practices, triggered by food security and climate change, have evolved precision farming. By combining machine learning with the Internet of Things (IoT), precision agriculture can optimize resource use and crop yields. Data analytics based on IoT are used here to enhance agricultural decision-making by using regression techniques like Support Vector Machines (SVMs) and Multilayer Perceptrons (MLPs). Data from environmental and soil sensors are collected in agricultural fields, including temperature, humidity, nitrogen, phosphorus, potassium, pH, and rainfall. Utilizing machine learning algorithms, this data is processed to predict which crops will yield the highest yields and utilize the most resources. Based on a comparative analysis, MLP models exhibit superior performance to SVM models with respect to training time, testing time, and regression error (lower RMSE). It achieves the highest classification accuracy (92%) among existing models such as SPS, CSMS, and SRIM. When farmers have access to real-time, data-driven insights, they can make better decisions, increase productivity, and adopt more sustainable farming practices.