本研究主要針對單期隨機性需求，探討多具時效性商品之最適訂購策略，因顧客到達的時間屬不確定性，所以將商品的銷售量視為一隨機變數，且為符合實務運作之考量，本研究將傳統單一商品擴充為多商品，並將具時效性商品的銷售狀況，劃分為平日與假日兩個部分進行探討，並建構含有空間與資金限制之報童模式，且利用本文提出的EM演算法求解，以達總期望利潤最大化之目標。接著，進行各相關參數對總期望利潤之敏感度分析，以進一步了解相關參數對總期望利潤之影響，且以一數值範例闡述本研究主題的意義，並列出六點具體結論以供後續研究與實務應用之參考。

The main propose of this article is to establish optimal ordering strategy for single period multiple perishable products under stochastic demand. Because the customer arrival time is Uncertainty, so we regard the product's sales quantity as a random variable, and then extend traditional newsboy problem. For practical applications, we divided sales period into weekday and holiday two conditions to explore, and construct multi-product newsboy problem model with space and budget constraints. However, an EM algorithm is provided in this paper that can maximize the total expected profit under space and budget constraints. Then, sensitive analysis is also taken to find out the influence of each parameter for total expected profit and ordering quantity. Then a numerical example is taken to illustrate the model and provide some insights. Finally six conclusions are drawn for practical applications and future studies.