SALES PREDICTION ON THE MARKET OF THE SELECTED GROUP OF GOODS AND SERVICES USING THE METHODS OF DATABASE MINING
The authors consider the trend models, which are based on mathematical dynamic graduation of actual sale values of the selected group of goods that took place in some time points, by selecting the functional relationship and calculation of its parameters. These trend models provide the ability to trend extrapolation of the identified constraints to the planned time points in order to obtain estimates of forecast sales. For the analysis of wealth information and diverse nature the authors study modern methods of data mining, which are the basis of Data Mining technology, and it allows finding significant correlations and relationships as a result of screening wealth information using modern methods of pattern recognition and application of unique analytical technologies.
Keywords: sales forecasting, functionally dependence, integrated approach, method of extrapolation, trend models, method of database mining, clustering, classification, regression, distribution of predicted values.
Olga A. Lukinova, Candidate of Technical Sciences, Associate professor. E-mail: lu_5555@mail.ru; Natalia D. Pisarenko, Candidate of Technical Sciences. E-mail: lu_5555@mail.ru; Lilia V. Smarchkova, Candidate of Economics, Associate Professor. E-mail: lilija-sma@rambler.ru. - Voronezh branch of the Russian Eco-nomic University named after G.V. Plekhanov; Pavel V. Samoilov, Candidate of Economics, Director of Voronezh College of Food and Processing Industry.