Development of a universal methodology for estimating the cost of educational services of Russian universities based on machine learning and mathematical modeling methods

This study is devoted to the development of a universal methodology for estimating the cost of educational services for higher education programs in Russian universities for 2026. The relevance of the study is due to fundamental changes in the regulatory framework for paid admission: the introduction of state quotas for places, the prohibition of setting prices below cost, and limiting price increases to the inflation rate. In conditions of peak demographic demand and the need to compensate for the loss of university income, there is a need for an objective tool for calculating the fair cost of tuition. The purpose of the work is to create a methodology suitable for practical use by universities. Using machine learning methods (gradient boosting, random forest), a systematic analysis of 30 pricing factors was carried out, and their normalized weighting coefficients were obtained. On this basis, mathematical models have been built that make it possible to calculate the reference cost of an educational program, assess its price adequacy, and predict price dynamics taking into account competitive pressures and macroeconomic factors. The developed methodology was tested using the example of a number of specialties at Samara University.