The rapid evolution of digital advertising demands intelligent systems capable of real-time learning and adaptive decision-making. This study proposes a Hybrid Reinforcement Learning–Optimization (RL–GA) framework that integrates reinforcement learning’s sequential adaptability with the global search and parameter-tuning capabilities of genetic algorithms. The hybrid architecture is designed to enhance ad-targeting accuracy, stability, and scalability in dynamic market environments. Empirical evaluation using real-world ad-interaction data demonstrates that the framework achieves superior targeting precision, faster convergence, and improved adaptability compared to conventional rule-based, GA-only, and standalone RL systems. The genetic optimization component enables continuous policy evolution, balancing exploration and exploitation, while reinforcement learning captures behavioral patterns across temporal and contextual dimensions. Qualitative analyses reveal that the model autonomously reallocates ad impressions toward high-engagement user segments, reflecting emergent contextual intelligence. The results affirm that the hybrid RL–GA framework provides a robust and data-efficient approach for adaptive advertising, establishing a pathway toward self-optimizing, behavior-aware marketing systems. This research contributes theoretically to hybrid intelligence and multi-objective optimization literature and offers practical insights for developing scalable, ethical, and transparent AI-driven advertising platforms.
Mahshid Asadollahi2 Mohammad Akbari Asl1* (Thu,) studied this question.