Advanced feature selection to study the internationalization strategy of enterprises

Area of Science:

• Business and Economics
• Computational Intelligence

Background:

• Firms operate in a complex global economic environment requiring international network access.
• Understanding internationalization determinants is vital for corporate success.
• Foreign Direct Investment (FDI) strategies are critical for market expansion.

Purpose of the Study:

• To apply and compare cutting-edge feature selection methods for enhanced understanding of FDI strategies.
• To gain deep knowledge about the key determinants driving corporate internationalization.
• To validate an intelligent system using real-world data from Spanish Multinational Enterprises (MNEs).

Main Methods:

• Evolutionary feature selection using a wrapper approach.
• Utilized Bagged Trees and Extreme Learning Machines as fitness functions for classification.
• Validated the intelligent system on empirical data from Spanish MNEs.

Main Results:

• Identified key features influencing the internationalization of Spanish MNEs.
• Demonstrated the effectiveness of the proposed intelligent system in uncovering internationalization drivers.
• Provided novel insights into corporate FDI strategies through advanced computational methods.

Conclusions:

• The intelligent system successfully identified critical factors for internationalization.
• This study offers a new computational approach to analyzing internationalization data.
• Findings provide valuable knowledge for firms seeking to expand internationally.