Research On Online Ordering Scheme For Extended Newsvendor Problem

Classic newsvendor problem is a kind of single-period problem in the research of inventory control,which has strong practicality and is widely used in inventory management and stock market environments.Its traditional solution is to provide ordering scheme based on the statistical assumption of demand information.However,the future demand in reality is often irregular and difficult to describe with appropriate probability distribution.In recent years,Weak Aggregating Algorithm(WAA)developed in theoretical computer provides a new method to solve the management decision problem which is completely out of statistical assumption.WAA is an online sequence prediction algorithm that integrates expert-group schemes.It only relies on the historical demand to provide the ordering scheme,which is completely out of statistical assumption of demand and can be applied to the solution of multi-period newsvendor problem.Thus,this thesis mainly applies WAA to study multi-period newsvendor problem completely out of statistical assumption and gives corresponding online ordering scheme for extended newsvendor problem.The main work of this paper includes the following three aspects.Firstly,WAA is applied to discuss the discrete multi-period newsvendor problem,and then salvage value and shortage price factors are taken into account to carry out the extended research.The specific online ordering schemes and their competitive performance analyses are given respectively.The effectiveness and robustness of the schemes are tested at last.Secondly,this thesis discusses the multi-period newsvendor problem under the price discount.It studies how to determine optimal ordering quantity to optimize certain target under the price discount mechanism.Online ordering schemes under continuous and discrete demand are given respectively and we further discuss all-unit and incremental-price discounts in the discrete demand case.Lastly,this thesis concentrates on the multi-period newsvendor problem with free-shipping factor.It provides scientific decision-making solution for newsvendor who order from Internet and have the opportunity to enjoy 0 shipping fee.Numerical examples are given to explore the influence of different parameters on the competitive performance of the provided schemes.The main innovations of this thesis are as follows:(1)It applies WAA that integrating expert-group schemes to the multi-period newsvendor problem and provide online ordering scheme for the discrete multi-period newsvendor problem.Besides,a loss function is constructed to get scientific online scheme for decision maker who is concerned about the performance gap with peer.This method completely out of statistical hypothesis of demand,which is an innovation in the ordering scheme research for multi-period newsvendor problem.(2)Considering the importance of online shopping and commercial promotions in practice,this thesis establishes and analyses mathematical model that considering the factors such as price discount,free-shipping and shortage cost.To enrich the researches on ordering schemes and their theoretical competitive performances,which make the model more suitable for actual decision-making.This has great significance to perfect the study of online newsvendor problem.