PREDICTIVE POWER UNLEASHED: A COMPARATIVE STUDY OF MULTIPLE LINEAR REGRESSION AND ARTIFICIAL NEURAL NETWORKS FOR NEW BUSINESS REGISTRATIONS IN THE DRC

Abstract

The level of entrepreneurship in a country is a critical indicator of economic vitality and growth. The World Bank Group introduced the "new business registered" indicator in 2006 to gauge entrepreneurship by considering various factors influencing business creation. Although research on this subject is limited, it holds significant importance, particularly for developing nations seeking economic enhancement. Identifying the factors impacting entrepreneurship and comprehending their interplay facilitates informed policy decisions for these countries. Prediction is a fundamental aspect of organizational decision-making, guiding responses to internal challenges. Selecting the appropriate prediction model for a specific case is a formidable challenge, influenced by numerous factors including the context, data availability, desired precision, time frame, cost-effectiveness, and analysis complexity. Information and communication technologies, especially decision support systems, offer avenues for precise predictions even in complex datasets. As The Democratic Republic of the Congo aspires to become an emergent nation by 2030, promoting entrepreneurship is pivotal. The "new business registered" indicator is paramount for assessing growth in this sector. Leveraging ICT for accurate prediction of this metric empowers policymakers to evaluate the effectiveness of current policies. This paper delves into the challenges of selecting the right model for decision support systems and its seamless integration into workflows, particularly in the context of enhancing entrepreneurship.