主讲人：Prof. Dr. Stefan Minner

讲座时间：2017年3月23日下午14：00

讲座地点：九里校区逸夫楼4304B

讲座内容简介：

In inventory management, demand forecasting and stock optimization are typically conducted sequentially. The data-driven approach suggests integrating both problems by optimizing inventory decisions based on historical data using mixed-integer programming. Thereby, forecast errors are penalized with their operational consequences. Further, the availability of large amounts of detailed data on a customer basis allows for using enhanced demand models in inventory theory.

The presentation reviews existing approaches for optimizing target inventory functions and safety stocks for several standard inventory problems including perishable products, dual sourcing, multi-echelon inventory systems, and identifying inventory replenishment patterns. Different exact and heuristics solution approaches to solve the integrated data-driven inventory problems will be presented. Practical applications from the retail sector illustrate the capability over traditional sequential approaches.

主讲人简介：

work field:现任慕尼黑工业大学，管理学院，物流与供应链管理，教授；曾任维也纳大学，工商管理学院，经济学与统计学、物流与供应链管理，教授；曼海姆大学，曼海姆商学院，工商管理与物流，教授；帕德博恩大学，经济学院，企业管理与企业计算、运营管理，教授；

Main competences:国际库存研究协会（International Society for Inventory Research，ISIR）主席, since 08/2010 ；欧洲物流协会研究发展委员会委员；德国物流协会专家委员会委员；德国运筹学会咨询委员会委员；编辑活动有： 《Logistics Research》杂志执行主编；《OR Spectrum》杂志（SCI，影响因子1.090）主编；《OR Spectrum》、《International Journal of Production Economics》、《Review of Managerial Science》 ,《Asia-Pacific Journal of Operational Research》编委；Management Science, Operations Research, Logistics, Supply Chain Management, Operations Management, Business Administration, and Applied Mathematics等主流期刊的评审专家。研究重点为全球供应链的设计与优化。研究工作包括开发决策支持工具，在零售和服务备件物流的研究和应用战略物流网络设计方法不确定性条件下的库存管理的具体应用。

学术海报：Voting System Model for Demand Forecast

主讲人：曹德弼教授

讲座时间：2017年3月23日下午14：00

讲座地点：九里校区逸夫楼4304B

讲座内容简介：

Prediction Market is one important method to do forecast.  IEM made great success last decades on electronic parts price forecast, presidential election, and so on. The method is applied in financial market frequently and given good performance.Now,the method is modified and used to predict demand for those so-called product the innovative product. One important point in the demand forecast for the innovative product is we know how the demand is uncertain and how we figure it out the risk of over production. In other words, certainly we need possible average future demand as well as the accurate volatility of the future demand.Traditional method rely on time series analysis, regulations, also deep learning recently a central concept of the AI,however any of them can not include implicit knowledge that owned by line managers, merchandisers or professional buyers. We developed a voting system which can integrate quantitative analysis and qualitative analysis so as to create“forecast information” from two types of data.

主讲人简介：

work field:生产与运作管理，物流与供应链管理，制造战略等。

Main competence:庆应义塾大学，理工学部（管理工学科）教授；东京工业大学                        特任教授（兼）；日本新华侨华人会副会长；学会及其他任职：曾任日本经营工学会                                  理事、监事；现任日本管理研修中心理事长；ISO/IEC JTC1/SC31　RFID标准化委员会委员长；日本国土交通省港湾IT战略委员会委员；亚洲管理科学与应用学会理事（发起人）；日本华人管理科学学会发起人；日本运作管理与战略学会理事；学术刊物任职：曾任日本经营工学会誌副主编；现任学术刊物Innovation and Supply Chain Management首席编辑；International Journal of Production Economics（SCI二区）编辑；Asian Journal of Management Science and Application 编辑；Information编辑。