The growing complexity of modern agriculture, coupled with challenges such as climate variability, pest outbreaks, and resource scarcity, demands intelligent and sustainable solutions. This paper proposes a Smart Agriculture Decision System (SADS) that integrates bio-inspired optimization algorithms, machine learning (ML), deep learning (DL), and IoT-enabled monitoring into a unified framework. The system is validated on crop yield prediction datasets, rice leaf disease datasets, and IoT-based smart farming datasets. Experimental results demonstrate that the integration of genetic algorithms (GA), particle swarm optimization (PSO), and moth flame optimization (MFO) significantly enhances system performance. The proposed framework improves yield forecasting accuracy, enables early disease detection, and optimizes water and fertilizer usage. The integrated approach advances precision agriculture and supports sustainable farming practices for both smallholder and large-scale farms.

Key Words: Smart Agriculture, Bio-Inspired Algorithms, Machine Learning, IoT, Yield Prediction, Disease Detection