| Summary: | Background. The problem of increasing the efficiency of forecasting sales in
the market for endoprosthetics of large human joints is considered, the solution of which is
associated with the need to take into account a large number of quantitative and qualitative
parameters depending on the choice of individual types of artificial joints and prosthetics
technologies. The conditions are shown under which the traditional approach to forecasting,
based on simple extrapolation and the method of scenarios, does not allow taking into account
the features of the product and the requirements of the target market. Materials and
methods. The information base was the results of scientific and applied research of specialists
in the field of forecasting the markets of medical devices, statistical data analysis and
machine learning. The paper shows the directions of using the methods of classification and
statistical training to predict the volume of sales of artificial human joints. Results. The
conditions for the application of regression trees models and logistic regression from the
standpoint of improving the forecasting accuracy in the target market are considered. The
possibility of using these models for solving inverse problems of finding the optimal set of
signs for classifying market segments and studying competition between products from the
standpoint of substitution or addition processes is shown. Conclusions. The results of the
study will make it possible to more effectively solve applied problems of planning the creation
of new products and assessing the potential for increasing sales in the endoprosthetics
market. They can be used to select the best ways to position and promote products on the
market.
|
|---|