방법 비교

	뉴스벤더 모형	안전 재고 및 재주문점 모델	확률적 최적화
분야	경영과학	경영과학	최적화
계열	Regression model	Regression model	Process / pipeline
기원 연도	1951	1998	1951 (SGD); 2014 (Adam)
창시자	Arrow, Harris & Marschak	Silver, Pyke & Peterson	—
유형	Stochastic single-period inventory optimization	Stochastic inventory control model	Gradient-based iterative optimization
원전	Arrow, K. J., Harris, T., & Marschak, J. (1951). Optimal inventory policy. Econometrica, 19(3), 250–272.	Silver, E. A., Pyke, D. F., & Peterson, R. (1998). Inventory Management and Production Planning and Scheduling (3rd ed.). Wiley. ISBN: 978-0-471-11947-0	Robbins, H. & Monro, S. (1951). A Stochastic Approximation Method. Annals of Mathematical Statistics, 22(3), 400-407.
별칭	Newsboy Model, Single-Period Inventory Model, Christmas Tree Problem, Gazete Satıcısı Modeli	Buffer Stock, Reserve Stock, Reorder-Point Model, Emniyet Stoğu	Stokastik Optimizasyon (SGD & Varyantları), stochastic gradient descent, SGD, Adam
관련	3	3	3
요약	The Newsvendor Model is a single-period stochastic inventory optimization framework that determines the profit-maximizing order quantity when demand is uncertain and unsold units cannot be carried forward. Formally introduced by Arrow, Harris, and Marschak (1951) in their foundational work on optimal inventory policy, the model balances the cost of ordering too much (overage) against the cost of ordering too little (underage) to yield a closed-form optimality condition known as the critical ratio.	Safety stock is an additional quantity of inventory held beyond expected demand during a replenishment lead time, designed to protect against stockouts caused by demand or supply uncertainty. Reorder-point models formalize this buffer by setting a trigger inventory level at which a new order is placed. Systematically developed within the stochastic inventory-control framework by Silver, Pyke, and Peterson (1998), the approach translates a desired customer-service level into a precise buffer quantity using the statistics of demand and lead-time variability.	Stochastic optimization is a family of iterative methods that minimize an objective function by computing gradients on randomly sampled subsets of data — mini-batches — rather than on the entire dataset at once. Pioneered by Robbins and Monro in 1951 as stochastic approximation, the approach became the standard engine for training large-scale machine-learning models through variants such as SGD with momentum, AdaGrad, RMSProp, and Adam.