Research on Computational Intelligence-based Workload Prediction in Edge Computing focuses on leveraging techniques such as neural networks, fuzzy logic, evolutionary algorithms, and hybrid AI methods to accurately forecast workloads and resource demands at edge nodes. This area addresses challenges including dynamic and heterogeneous workloads, resource constraints, limited connectivity, and the need for real-time prediction to optimize task scheduling, resource allocation, and energy management. Key research directions include time-series forecasting using deep learning and recurrent neural networks, hybrid intelligence models combining machine learning and optimization techniques, and context- and mobility-aware workload prediction for IoT and mobile applications. Other emerging topics involve adaptive and online learning methods for changing edge environments, collaborative edge–cloud prediction frameworks, and predictive resource scaling to minimize latency, energy consumption, and operational cost. Additionally, research on fault-tolerant and privacy-preserving workload prediction represents significant avenues for advancing efficient, intelligent, and proactive edge computing systems.