Numerous technoeconomic factors have had a considerable impact on the industrial and service sectors. The best uses of production and service resources have been a main research area. Capacity planning, which aims at the optimum utilization of resources, has attracted a lot of scholarly attention. To the best of our knowledge, however, there has never been a scientometric analysis published on this subject; as a result, the goal of this article is to provide the first such analysis. The goal of this study is to gain a thorough structural view of the background and important facets of capacity expansion and planning research. The Web of Science database, VOSviewer, and CiteSpace software are used to give an overview. Our research indicates that after the year 2000, the publication rate increased significantly. Applications for capacity planning can be divided into eight main groups based on the authors' keyword analysis. We also list both well-known and underutilized modeling and solution methods. Recent applications, research trends, and some omissions are highlighted in this study. Leading contributing countries, institutions, and the most common research outlets are mapped and analyzed. Finally, some potential future research topics are suggested

The evolving landscape of cybersecurity threats, characterized by increasingly sophisticated and adaptive attackers, poses major challenges to traditional static network defense mechanisms. To address these limitations, this paper proposes an adaptive cyber defense framework that integrates Reinforcement Learning (RL) with Attack Graph (AG) modeling. The interaction between attacker and defender is formulated as a repeated zero-sum stochastic game over a partially observable Attack Graph-guided environment, allowing both agents to adapt their strategies through repeated interaction. Two value-based learning approaches are investigated, namely tabular Q-learning and Deep Q-Networks (DQN), under a unified attacker–defender setting. Experimental results across multiple training scenarios show that defender performance improves substantially as the training budget increases. Under limited training, Q-learning provides a computationally efficient and stable baseline, while DQN requires more training and careful tuning to achieve strong performance. However, with extended training, the DQN-based defender attains the highest win rate, albeit at a significantly greater computational cost. In addition, multi-run statistical comparisons highlight a clear trade-off between defensive effectiveness and runtime efficiency: Q-learning remains far more lightweight, whereas DQN offers stronger asymptotic performance when sufficient resources are available. These findings demonstrate the promise of learning-based adaptive defense over attack graphs while also emphasizing the importance of training budget, computational constraints, and model selection in practical cyber defense deployment.