Pages: 7-20
Introduction. The present study focused on investigating entrepreneurial intention in Romania using a technique specific to the field of data mining, respectively classification trees. The main idea of the study is to identify the probabilities of an individual being classified as a potential entrepreneur, having as a starting point a series of socio-demographic characteristics, but also perceptions, beliefs and opinions on issues related to training, experience and skills.
Aim of the study. The analysis takes into account some perceptions of the institutional support that authorities can provide to the entrepreneurial ecosystem. After analyzing 1522 responses to a questionnaire, it was concluded on the one hand that most respondents are not attracted to the idea of launching their own businesses, while potential entrepreneurs are characterized by a high confidence in their skills, in turn know individuals who have become entrepreneurs, these evidence being in line with other studies in the field, these aspects characterizing other groups of individuals and other societies. The low importance of the variables that describe age, gender and fear of failure in building the model implies that in Romania these characteristics do not significantly influence the decision of individuals to become entrepreneurs.
Keywords: entrepreneurial intention, classification trees, GEM database, self-confidence
JEL Classification: C38, C82, L26, M13
