在不对需求做任何统计假设的情形下,该文用理论计算科学兴起的集成专家意见的弱集成算法研究多阶段报童决策。弱集成算法是一种指数加权平均集成方法,在一定的初始权重下,根据损失函数在线调整专家意见的权重。基于收益损失函数和固定订购量的专家意见,得到了与从累积收益角度研究相一致的决策方法;并扩展研究了带有回收价值的情形。理论上证明了决策方法的累积收益损失几乎不超过最优专家意见的累积收益损失。通过数值算例验证了决策方法的可行性和合理性,探讨了卖出价和成本价等因素对竞争性能的影响,说明了回收价值的引入大大提高了决策方法的竞争性能,具有重要的现实意义。
The Weak Aggregating Algorithm (WAA) of prediction with expert advices, which advanced in computer science, is applied to study the multi-period newsvendor problem without making statistical assumption. WAA is an exponentially weighted average algorithm that updates the expert advice's weight according to loss function with initial weights distribution. Based on the return loss function and the expert advice of fixed stock level strategy, the decision-making method is used in this paper, which is in accord with the conclusions obtained using return function; and the case with salvage value is extended. Theoretically, it is proved that the cumulative loss the proposed decision-making method achieved does exceed that of the best expert advice. Numerical examples are presented to further illustrate the feasibility and rationality of the proposed decision-making method and explore the effect of selling and cost price on competitive performance;the results show that the introduction of salvage value greatly improves the competitive performance of the proposed decision-making method and thus presents important practical significance.
[1] Petruzzi N C, Dada M. Pricing and newsvendor problem: A review with extension[J]. Operations Research, 1999, 47( 2) : 183-149.
[2] Silver E A, Pyke D F, Peterson R P. Inventory management and production planning and scheduling[M]. New York: John Wiley, 1998.
[3] Khouja M. The single-period (news-vendor) problem: Literature review and suggestions for future research[J]. Omega, 1999, 27(5): 537-553.
[4] 汪小京, 刘志学, 郑长征. 多类顾客环境下报童模型中库存分配策略研究[J]. 中国管理科学, 2010, 18(4): 65-72.
[5] 黄松, 杨超, 张曦. 考虑战略顾客行为带预算约束的多产品报童问题[J]. 中国管理科学, 2011, 19(3): 70-78.
[6] 许民利, 李展. 基于CVaR准则具有预算约束和损失约束的报童决策[J]. 控制与决策, 2013, 28(11): 1614-1622.
[7] 周艳菊,应仁仁,陈晓红,等. 基于前景理论的两产品报童的订货模型[J]. 管理科学学报, 2013, 16(11): 17-29.
[8] Scarf H. Some remarks on Bayes solutions to the inventory problem[J]. Naval Research Logistics Quarterly, 1960, 7(4): 591-596.
[9] Scarf H. Bayes solution to the statistical inventory problem[J]. Annals of Mathematical Statistics, 1959, 30(2): 490-508.
[10] Karlin S. Dynamic inventory policy with varying stochastic demands[J]. Management Science, 1960, 6(3): 231-258.
[11] Iglehart D L. The dynamic inventory problem with unknown demand distribution[J]. Management Science, 1964, 10(3): 429-440.
[12] Liyanage L H, Shanthikumar J G. A practical inventory control policy using operational statistics[J]. Operations Research Letters, 2005, 33(4): 341-348.
[13] O'Neil S, Chaudhary A. Comparing online learning algorithms to stochastic approaches for the multi-period newsvendor problem[C]. Proceedings of the Tenth Workshop on Algorithm Engineering and Experiments, San Francisco,California,January19,2008.
[14] Huh W T, Janakiraman G, Muckstadt J A, et al. An adaptive algorithm for finding the optimal base-stock policy in lost sales inventory systems with censored demand[J]. Mathematics of Operations Research, 2009, 34 (2): 397-416.
[15] Huh W T, Levi R, Rusmevichientong P, et al. Adaptive data-driven inventory control policies based on Kaplan-Meier estimator for censored demand[J]. Operations Research, 2011, 59(4): 929-941.
[16] Huh W T, Rusmevichientong P. A non-parametric asymptotic analysis of inventory planning with censored demand[J]. Mathematics of Operations Research, 2009, 34 (1): 103-123.
[17] Zhu Zhisu, Zhang Jiawei, Ye Yinyu. Newsvendor optimization with limited distribution information[J]. Optimization Methods and Software, 2013, 28(3): 640-667.
[18] Kwon K, Cheong T. A minimax distribution-free procedure for a newsvendor problem with free shipping[J]. European Journal of Operational Research, 2014, 232(1): 234-240.
[19] Sleator D, Tarjan R. Amortized efficiency of list update and paging rules[J]. Communications of the ACM, 1985, 28(2): 202-208.
[20] Karlin A R, Manasse M S, Rudolph L, et al. Competitive snoopy caching[J]. Algorithmica, 1988, 3(1): 79-119.
[21] Borodin A, El-Yaniv R. Online computation and competitive analysis[M]. Cambridge:Cambridge University Press, 1998.
[22] Wagner M R. Fully distribution-free profit maximization: the inventory management case[J]. Mathematics of Operations Research, 2010, 35 (4): 728-741.
[23] Wagner M R. Online lot-sizing problems with ordering, holding and shortage costs[J]. Operations Research Letters, 2011, 39(2): 144-149.
[24] 张桂清,徐寅峰. 报童问题的最优竞争比策略及其风险补偿模型[J]. 管理学报, 2011, 8(1): 97-102.
[25] 张桂清,徐寅峰. 概率预期下在线报童问题的最小风险策略[J]. 中国管理科学, 2010, 18(6): 131-137.
[26] Ball M, Queyranne M. Toward robust revenue management: Competitive analysis of online booking[J]. Operations Research, 2009, 57 (4): 950-963.
[27] Van den Heuvel W, Wagelmans A P M. Worst case analysis for a general class of on-line lot-sizing heuristics[J]. Operations Research, 2010, 58 (1): 59-67.
[28] Cesa-Bianchi N, Lugosi G. Prediction, learning, and games[M]. Cambridge:Cambridge University Press, 2006.
[29] Vovk V. Competitive on-line statistics[J]. International Statistical Review, 2001, 69(2): 213-248.
[30] Kalnishkan Y, Vyugin M V. The weak aggregating algorithm and weak mixability[J]. The Journal of Computer and System Sciences, 2008, 74(8): 1228-1244.
[31] Levina T, Levin Y, McGill J, et al. Weak aggregating algorithm for the distribution-free perishable inventory problem[J]. Operations Research Letters, 2010, 38(6): 516-521.
