Формування структури вхідних даних для базового прогнозу вартості вживаних автомобілів

Abstract

The paper considers the features of forming the input data structure for a baseline forecast of used car value. The feasibility of using market data from online listings, vehicle technical characteristics, and their combination within a combined approach is analyzed. It is determined that market listings make it possible to take into account the current situation in the used car market, but require cleaning and relevance checking. Technical characteristics allow a more accurate description of a specific vehicle, but without the market component they do not fully reflect the current price situation. It is substantiated that the most appropriate solution is the use of a combined data structure that includes both market and technical parameters and can be used as a basis for further value forecasting.