大型小売店における売上予測(第3報)

Abstract

A sales forecasting is very important for making a management plan and proceeding development and product planning. There are various forecasts such as technical forecast, economic forecast, sales forecast and product forecast. Succeeding to the previous report, the present study discusess further sales forecast of clothing in the department stores in Aichi Prefecture by means of a multiple regression analysis. In the previous report, such dependent variables were used as consumer expense, sales floor area, time and dummy variables. We know well that department's sales are influenced by temperature. Then the present study added a temperature variable to the dependent variables. As the result of analysis, the autocorrelation of the residual are improved very much in which the forecast error rates are 6.86%max. and -1.57%min.