Genetic programming based machine learning in classifying public-private partnerships investor intention / Ahmad Amin ... [et al.]

To accelerate the growth of public infrastructure development, the government employs public private partnerships (PPP). However, this scheme exposes the private sector to various risks, including political risks, which can negatively impact the financial performance and reporting of participating firms. A significant challenge for the government is the insufficient private sector engagement in PPP arrangements. Hence, the purpose of this study is to evaluate the effectiveness of machine learning prediction models in categorizing private investor interest in PPP programs based on Indonesia evidences. The PPP data was analyzed in this study using two machine learning approaches, Genetic Programming and conventional machine learning, with testing results showing that all machine learning algorithms from both approaches achieved high accuracy rates of over 80%, with the Genetic Programming machine learning outperformed the conventional approach. This study highlights the potential of machine learning algorithms in predicting private investor interest in PPP programs, providing a tool for managing political risks and encouraging greater private sector participation.