Academy of Entrepreneurship Journal (Print ISSN: 1087-9595; Online ISSN: 1528-2686)

Abstract

Comparative Analysis of A Traditional and Machine Learning Techniques in Predicting SMMES Growth Performance

Author(s): Helper Zhou and Victor Gumbo

Firm growth remains one of the most important subjects in research today. This mainly owes to the important role played by growth-oriented firms in addressing socio-economic challenges largely facing governments in developing countries. As such, this study aimed to identify and model key growth drivers of Small-Medium, and Micro Enterprises (SMMEs) harnessing traditional and emergent machine learning techniques. The study further compared the growth predictive modeling performance of the traditional logistic regression and two machine learning techniques, Artificial Neural Network (ANN) and Support Vector Machine (SVM) in predicting SMME growth. The study utilized three-year panel dataset from 191 SMMEs in the manufacturing sector in South Africa’s second-largest province of KwaZulu Natal. The results showed that the duo of SVM and ANN performed better than Logistic Regression in predicting firm growth. Sales revenue was identified as the most important driver of growth and it was recommended that key stakeholders can leverage this key driver to drive the sustainability of SMMEs. Overall, the study recommended the adoption of the SVM technique for SMME growth predictive modeling.

Get the App