AI-IoT Enabled Smart Agriculture Using Sequential Hybrid Feature Selection for Intelligent Irrigation Management

Efficient irrigation management is a prerequisite of sustainable agriculture, especially when considering water constraints and climate variability. Recent breakthroughs in Artificial Intelligence (AI) and Internet of Things (IoT) technology have made it possible to develop smart agricultural systems that monitor environmental conditions and automate irrigation decision-making. However, high-dimensional agricultural sensor data usually contain redundant and unnecessary features, which significantly impair machine learning model performance. This paper presents a smart agriculture system based on AI-IoT with an intelligent irrigation control using a Sequential Hybrid Feature Selection (SHFS) method. The SHFS model sequentially combines Principal Component Analysis (PCA), Particle Swarm Optimization (PSO) and L1-regularized Logistic Regression (Lasso) to minimize the dimensionality and preserve the discriminative information. The suggested framework was evaluated on 100,000 IoT-based agricultural observations consisting of ambient as well as soil-related factors. We evaluated the prediction capability of some classifiers, such as k-Nearest Neighbors (kNN), Na¨ıve Bayes, Logistic Regression and XGBoost. The experimental results show that the SHFS framework can achieve a significant improvement in classification performance compared to traditional feature selection methods. The best accuracy was 97.5%, attained by XGBoost in combination with SHFS, which also had higher precision, recall, and F1-score values than the other models. The collected findings demonstrate that the suggested framework efficiently reduces data redundancy, enhances the reliability of irrigation forecasts, and promotes sustainable applications of precision agriculture.