Модификация модели прогнозирования финансово-экономического состояния для задачи управления развитием медицинских учреждений

Abstract

The article analyzes the model of forecasting the financial and economic condition, considers the possibility of its transformation for managing the development of medical institutions, and provides an appropriate modification using methods of discriminant analysis and mathematical programming. The formulated modification of the model has a clear mathematical justification, which allows, on the one hand, to prove its validity, and on the other hand constructively shows the possibility of its use, taking into account the specifics of the medical industry. The purpose of the work. The purpose of this work is to study changes in the forecasting model of the financial and economic condition, taking into account the specifics of the industry, as well as to evaluate the application of methods of discriminant analysis, reducing them to a mathematical programming problem, the limitations of which are reduced to systems of linear inequalities obtained by building maximally joint subsystems. Materials and methods. It is proposed to use the method of discriminant analysis with extensible precedent sets. It is assumed that in order to improve the quality of solving this problem, expert judgments about objects that were not initially included in the training sample are involved. At the same time, the goal is to determine the best objects to involve them in the learning material in order to maximize the solution of the problem. To formalize the model based on the predicted values, we will encrypt all available data into a vector of states. The considered methods show their effectiveness. The results of the study. During the modification of the economic and financial condition model, a conclusion is made about the linearization of the model, which is also associated with the differentiation operation. The possibility of using neural networks is also demonstrated for the case of differentiation of complex functions, and in the case of convex programming, for marginal estimates obtained using the directional differentiation operation. Conclusion. The results obtained allow us to assert that the use of strict mathematical tools, a modified model of the economic and financial condition for the medical industry is easy to use for development management, since it can be implemented as some component of a computer program (neural network) into which the administrator enters data, and at the output receives a processed result.