Artificial Hummingbird Algorithm (AHA) is a nature-inspired metaheuristic optimization technique that mimics hummingbird flight behaviors, foraging strategies, and memory mechanisms. AHA has garnered growing interest in computational intelligence and engineering applications because of its population-based search capacity and adaptable exploration–exploitation processes. The biological inspiration, mathematical formulation, guided foraging, territorial foraging, migration foraging, axial, diagonal, and omnidirectional flying patterns, and visit-table-based memory mechanism of AHA are all thoroughly reviewed in this study.  The use of AHA in classification issues, where it can aid in feature selection, hyperparameter optimization, and the creation of hybrid learning frameworks, is given special attention. To determine the possible contributions of AHA to classification performance and computing efficiency, representative application fields such as biomedical signal analysis, image identification, defect detection, voice processing, and power quality disturbance classification are investigated. Additionally, convergence behavior, local-optimum avoidance, computing complexity, parameter sensitivity, and generalization capabilities are examined in connection to the advantages and disadvantages of AHA. Benchmark selection, baseline comparisons, numerous categorization measures, and repeatable experimental conditions are among the methodological factors that the review emphasizes for equitable performance evaluation. Lastly, new avenues for research are described, including explainable classification, imbalanced datasets, adaptive and hybrid AHA variations, and real-world validation. This review lays the groundwork for future studies on trustworthy, effective, and robust intelligent classification systems by offering an organized understanding of AHA and its function in optimization-assisted classification.

Keywords: metaheuristic optimization; Artificial Hummingbird Algorithm; bio-inspired optimization; classification problems; hyperparameter optimization; feature selection; machine learning; computational intelligence.