Background and Objectives: The present study aimed to identify effective indicators and components in value chain smartization and predict the level of smartization in companies located in Caspian Industrial Park.
Materials and Methods: This study is applied in terms of purpose and quantitative in terms of nature and method. The reason for using this approach is that in the quantitative part, smartization and prediction of value chain performance are carried out using mathematical modeling and artificial neural networks. The research is also considered descriptive-analytical and developmental in terms of implementation method, because in addition to analyzing the current situation, it leads to the design of an intelligent algorithm.
Results: Data analysis using the artificial neural network algorithm showed that the model is able to predict the level of smartization of the value chain with high accuracy. The results of the model performance evaluation with MSE, RMSE, and coefficient of determination (R²) indices showed that the neural network was able to model the nonlinear and complex relationships between the input and output indices well. In addition, the sensitivity analysis of the input variables revealed that digital infrastructure and data-driven decision-making have the greatest impact on the level of smartization.
Conclusion: Using artificial neural networks allows for accurate prediction of the level of smartization, identification of effective factors, and prioritization of management actions. In addition to prediction accuracy, this method has high generalizability and can be used as a practical tool for planning and improving value chain smartization in industrial companies.