透過您的圖書館登入
IP:3.141.24.134
  • 學位論文

預測模型於零售商訂貨決策之應用

Ordering Decisions of Newsvendors with Forecasting Methods

指導教授 : 楊奕農

摘要


在面對市場需求的不確定及假設未來需求符合某一種隨機分配的情況下,以追求銷售利潤最大為目標,零售商訂貨理論可按照商品售價、進貨成本與殘值來推導一個常數值的最適訂貨量。然而,現實中未知的需求量往往並非是一隨機變數,而常常具有自我相關或季節性變動之性質,因此雖然零售商訂貨決策理論提供了如何訂貨以獲得最大利潤的基礎,但若能再透過各種預測模型進行零售需求預測,進而以此決定非固定常數之訂貨數量,對決策者而言,是不是比零售商訂貨理論所推估而得的訂貨量,可獲得更高的獲利,那成為本研究的重點。本研究利用某連鎖超商之北區某一分店,其中A、B二份報紙銷售量為研究樣本,以移動平均法、簡單指數平滑法、賀特的線性指數平滑法、溫特的三個參數指數平滑法及最小平方法做需求預測並與傳統零售商決策理論之訂貨獲利進行比較以決定訂貨量,計算及比較所得之預期毛利,本研究發現,在五種需求預測模型所推得之訂貨量中,除B報由報社以移動平均法、簡單指數平滑法、賀特的線性指數平滑法做需求預測所對應之利潤比傳統零售商訂貨理論的利潤差外,其它的皆優於傳統零售商決策理論的利潤,其中以含趨勢、星期虛擬變數及落後期之最小平方法做需求預測所對應之利潤為最高,而溫特的三個參數指數平滑法所對應的結果則次之。

並列摘要


According to Newsvendor theory, retailers can make a constant ordering quantity by sale price, purchase cost and product surplus value under the normal distribution demand. However, the demand usually is not normal distribution in the actual world, it often has the autocorrelation and season trend. Newsvendor theory provides a good foundation to get maximum profit. If we use the forecasting method to forecast the demand, and farther, use those demand to make a non-constant ordering quantity, it should be more useful than traditional newsvendor theory to increase the profit. It is the purpose of this paper. In this paper, we use the five ways to set the demand forecasting, the moving average method, the simple exponential smoothing method, Holt,s linear exponential smoothing method, Winters, three-parameter linear and seasonal exponential smoothing method, and ordinary least squares principle that applies on two different newspapers to decide the order quantity base on the newsvendor theory, then compare the effect of the profit. The result indicate the ordering decision with the forecasting base on the newsvendor theory can improve the decision maker’s profit than tradition newsvendor theory and compare the five forecasting methods of this paper, taking the method of ordinary least squares principle with the seasonal dummy variable have the best effect.

參考文獻


黃營芳、張志瓏 (2013),台灣飲料店業的銷售預測模型,商業現代化學刊,7 卷 2 期 ,頁227 - 242。
陳榮志 (2008),預測分析運用於訂購模式之研究-以A公司為例,成功大學工程管理系碩士論文。
游智元 (2006),時間序列方法於台鐵短期旅運需求預測之研究,成功大學工程管理系碩士論文。
羅慕君 (2004),短期訂單預測模型之研究-PDA產業為例,中原大學資訊管理學系碩士論文。
Aydın, G., & Porteus, E. L. (2008). Joint inventory and pricing decisions for an assortment. Operations Research, 56(5), pp.1247-1255.

延伸閱讀