The Model of Human Capital Management in Decision Support Systems

• Main Authors: O. D. Kazakov, N. Y. Azarenko
• Format: Article
• Language: Russian
• Published: Voronezh state university of engineering technologies 2019-12-01
• Series: Vestnik Voronežskogo Gosudarstvennogo Universiteta Inženernyh Tehnologij
• Subjects: human capital, capital development, digital economy, knowledge, machine learning, random forest
• Online Access: https://www.vestnik-vsuet.ru/vguit/article/view/2280

A system model of human capital management has been developed, which, according to the authors, can be an important addition to building an effective decision support system in the field of socio-economic development of the region. The interaction of the subjects of the model can occur in order to implement innovative processes in IT organizations. The developed system model of human capital management takes into account the peculiarities of human capital management in the digital economy and includes three subsystems: “The structure of human capital”, “Tools for the development of human capital”, “Organizational and economic tools for managing human capital”. The work used data obtained in the analysis of more than 100 innovative enterprises engaged in research, including in the field of information technology. As a result of the study, marked-up data were generated on the level of development efficiency of human capital based on the initial processing of financial statements of innovative organizations. To determine the directions of improving the quality of human capital, an approach to assessing the level of efficiency of its development based on the machine learning model is presented. The values ??of the quality metrics of the following machine learning algorithms for solving the problem are presented: linear regression; Random Forest; nearest neighbors method. To classify the region’s enterprises according to the level of development efficiency of human capital based on open financial information, the Random Forest algorithm was chosen. 11 most accurate classification rules in a hierarchical sequential structure were identified and formulated. This will allow for more complete consideration of all aspects of intellectual resources management in the digital economy.