The Application of Machine Learning for Creating a Typology of Universities' Financial Models

Abstract

This study presents an application of machine learning for creating a typology of Russian universities’ financial models. Large-scale national initiatives aimed at enhancing human potential and academic excellence, such as Project 5-100, university-industry consortia, world class research center programs as well as the Priority-2030 program, require relevant financial and management accounting tools enabling appropriate analyses of universities’ contribution to national scientific policy implementation. However, when conventional financial analysis and audit techniques are adopted from the corporate sector, they may prove to be irrelevant for assessing the societal impacts of universities. Existing impact study methods, such as those applied in the Russell Group universities’ impact assessment, are expensive and time consuming, so promising machine learning techniques and existing open data from government information systems were used in this study to assess universities’ financial models.

Keywords

• science and technology policy
• typology of universities
• machine learning