Measuring economic resilience of manufacturing organization leveraging integrated data envelopment analysis (DEA)-machine learning approach

Abstract

ABSTRACTThis study aims to assess the economic resilience of manufacturing firms through a combination of output-oriented data envelopment analysis (DEA) and machine learning techniques. The research draws on economic resilience factors identified in the literature and focuses on three categories: economic-related factors (financial flexibility, microeconomic market, macroeconomic stability), production-related factors (restoration of production, backup inventories, resource pooling/sharing), and management-related factors (diversification of activities, good governance (management), relocation). Using DEA, a mathematical approach, the study computes and analyzes the contributions of various components to economic resilience. The results of DEA normalization indicate that the highest weighted criteria are financial flexibility, good governance (management), and resource pooling (sharing). To gain a deeper understanding of the data structure, the K-means algorithm is employed for clustering and analysis. K-means clustering is a popular exploratory data analysis technique that aims to group samples into clusters of equal variances by minimizing inertia or the sum of squares within each cluster. The combination of these techniques with sensitivity analysis provides a novel analytical approach for policy formulation and decision-making. The findings have implications for practitioners and domain experts, offering valuable insights into enhancing economic resilience in the manufacturing sector.JEL CLASSIFICATION: PRISMAKEYWORDS: Economic resiliencedata envelopment analysis (DEA)machine learning approachdecision-making unit (DMU)K-means clustering Disclosure statementNo potential conflict of interest was reported by the author(s).