기상정보를 활용한 의류제품 판매예측 시스템 연구: S/S 시즌 제품을 중심으로

Abstract

This study aims to develop clothing sales forecast system using weather information. As the annual temperature variation affects changes in daily sales of seasonal clothes, sales period can be predicted growth, peak and decline period by changes of temperature. From this perspective, we analyzed the correlation between temperature and sales. Moving average method was applied in order to indicate long-term trend of temperature and sales changes. 7-day moving average temperature at the start/end points of the growth, peak, and decline period of S/S clothing sales was calculated as a reference temperature for sales forecast. According to the 2013 data analysis results, when 7-day moving average temperature value becomes <TEX>$4^{\circ}C$</TEX> or higher, the growth period of S/S clothing sales starts. The peak period of S/S clothing sales starts at <TEX>$17^{\circ}C$</TEX>, up to the highest temperature. When temperature drops below <TEX>$21^{\circ}C$</TEX> after the peak temperature, the decline period of S/S clothing sales is over. The reference temperature was applied to 2014 temperature data to forecast sales period. Through comparing the forecasted sales periods with the actual sales data, validity of the sales forecast system has been verified. Finally this study proposes 'clothing sales forecast system using weather information' as the method of clothing sales forecast.