Integrating support vector regression and metaheuristic algorithms for novel prediction in simulating intercity corporation networks

Intercity corporate networks have an important role in increasing enterprise efficiency and competitiveness. Therefore, they act as an engine for economic development and regional cooperation. The ability to predict and model them appropriately will lay the foundation for decision-making and optimization of the distribution of resources in such a way that effective communication across corporations is ensured. This paper presents an approach to the predictive modeling of a support vector regression to simulate the intercity corporation network. Precise data prediction and transmission across the network are very important for any simulated model to be deployed. Apart from optimization methods, the incorporation of meta -heuristic algorithms elevates the accuracy and speed of the forecast. This research investigates six optimization methods and their hybridization with SVR, with a critical investigation and comparison in terms of statistical performance metrics. It can be observed from the results that both the Manta-Ray Optimizer and Battle Royale Optimizer result in good performances with low error rates and high values of R and R 2 . In this regard, the Manta-Ray Optimizer is chosen as the final optimizer for the proposed hybrid algorithm since it had an R 2 value of 0.9430 in the test data, followed by the Salp Swarm Optimization Algorithm with an R 2 value of 0.9410, and the Battle Royale Optimizer with the lowest R 2 value of 0.9320 observed for the test data.