March 15, 2026
Ahmad Ghorbanpour

Ahmad Ghorbanpour

Academic Rank: Assistant professor
Address:
Degree: Ph.D in Industrial management
Phone: 09112919807
Faculty: School of Business and Economics

Research

Title Discriminant evaluation of circular economy performance in manufacturing industries using machine learning algorithm
Type Article
Keywords
Discriminant evaluation · Circular economy performance · Manufacturing industries · Machine learning
Journal PRODUCTION ENGINEERING-RESEARCH AND DEVELOPMENT
DOI
Researchers maryam bakhtar (First researcher) , Ahmad Ghorbanpour (Second researcher) , Khodakaram Salimifard (Third researcher)

Abstract

In recent years, companies have needed a new paradigm of economy for sustainable development to defend the environ- mental and sustainable values of society. The circular economy can serve as an efective tool to reduce environmental impacts and help organizations achieve better performance. Recently, several studies on the circular economy have been published, but few have examined the performance of manufacturing industries in terms of circular economy measures. The aim of this paper is to provide a discriminant evaluation of the circular economy performance of manufacturing industries. By reviewing the literature and interviewing experts, fourteen features of the circular economy were identifed. Additionally, Student’s t-test was used to select the features. In this study, the performance of manufacturing industries in adopting circular economy practices was analyzed, which should be distinguished from general sustainability, as circular economy focuses on resource optimization, waste reduction, recycling, and closed-loop production. The features selected for analysis included hiring and training of environmental staf, employees with specialized environmental expertise, vol- ume of recycled green space waste, volume of recycled paper and cardboard waste, volume of recycled glass waste, green energy consumption, electricity consumption, green purchases, raw materials consumed by parent companies, and raw materials purchased domestically. To group industries based on these features, the k-means algorithm was used, clustering similar industries together, and discriminant analysis was applied to validate the clusters and develop a predictive function for classifying industries into clusters based on their feature values. Cluster analysis revealed two distinct groups: Cluster1 performed well in gas consumption, green energy consumption, and investment in environmentally friendly technologies but below average in other features, while Cluster2 showed strong performance