Enhancing Export Competitiveness of SMEs in Sfax: A Machine Learning Approach Using Principal Component Analysis

The competitiveness of firms is a subject that is frequently discussed these days by managers, lawmakers, and academics. Even though the idea of competition may seem straightforward, it is frequently used in a variety of dubious contexts. Although the definition of competition might appear basic, there are many skeptical applications of this idea. In order to increase the position of small and medium‐sized exporting enterprises (SMEs) in Sfax, the economic capital of Tunisia, on the international market, our goal is to identify the key elements driving improvement in their competitiveness. On the basis of theoretical and empirical research in this field, we identified, based on state‐of‐the‐art recommendations, a set of 19 key criteria (factors) that are crucial for preserving an enterprise’s competitiveness in export. The data used in this study was collected using a questionnaire addressed to business leaders and then evaluated on a Likert scale by experts. Principal component analysis (PCA) modeling was used alongside machine learning algorithms to identify the relationships between these factors as well as to determine the factors capable of influencing the competitiveness of 40 firms. The initial number of variables in our data was 70; using PCA, we reduced this number to 27 for our first experiment and to 14 for our second experiment. Using data augmentation techniques provided by the Python programming language, we increased the number of firms to 60. We managed to achieve an F‐score of 74.76% by using the random forest algorithm through the application of PCA modeling for 14 features selected. On the other hand, the energy, chemistry, and rubber industry sector has the highest F‐score of 85.71% followed by the textile, clothing, and shoes industry with an F‐score of 80%. These findings provide valuable insights into the factors that can propel SMEs in Sfax toward global market competitiveness.