Providing a Smart Startup Model Based on Business Intelligence Capacity, Network Learning, and Innovation

Abstract

The aim of this study is to develop a smart startup model based on business intelligence capacity, network learning, and innovation. It is assumed that smart startups are influenced by these factors, and in the present study, the importance of these components has been determined based on their weighting according to startup productivity. The weighting of business intelligence system components and the indices of network learning and innovation was conducted using the Grey Relational Analysis (GRA) method with interval fuzzy numbers. In the next step of the study, a multi-level mathematical model incorporating smart startup components was designed as a multi-objective model based on the weights determined in the previous step. To solve this model, a multi-objective metaheuristic algorithm based on the Pareto archive was implemented. After determining the set of optimal solutions, the obtained solutions were analyzed, and the final solution was selected for optimizing the startup model. The objective function of the problem is to minimize the completion time of the scheduling program. This problem belongs to the class of NP-Hard problems in MATLAB software. In the proposed metaheuristic methods, a column generation technique was combined with the Genetic Algorithm (GA), the Biogeography-Based Optimization (BBO) algorithm, and a hybrid algorithm combining GA and BBO (Hybrid Heuristic Algorithm - HH). Computational results related to algorithm performance analysis indicate that the three metaheuristic algorithms GA, BBO, and Hybridize exhibit significant differences. Consequently, from a statistical perspective and with a 95% confidence level, the algorithm performances in terms of effectiveness rank as follows: GA in the first position, BBO in the second, and Hybridize in the third. The findings of the study suggest that GA is recognized as the superior modeling method in this context. This is valuable for the industry and management as it demonstrates that among the available tools for modeling smart startups, GA can deliver the best performance.