2016 INFORMS Annual Meeting Program

201 / 561

INFORMS Nashville – 2016
201
3 - An Inventory Model With Fuzzy Demand In a Two-echelon Supply
Chain Management: Drug Distribution In Hospital Pharmacy
Parisa Jannatifard, graduate Student and teacher assistant,
Southern Illinoise University Edwardsville, Edwardsville, IL,
62026, United States, , Javad Sadeghi
This paper develops an inventory model with fuzzy demand regarding the ven-
dor managed inventory (VMI) policy. In recent past few decades, VMI policies
have been used in modeling inventory problems. The vendor’s warehouse has a
capacity constraint while a vendor supplies several products to retailers. This pol-
icy reduces total costs in healthcare system like drug distribution in hospital
pharmacy. The aim of this paper is to find a near optimal solution including
order quantities for vendor and retailer and replenishment frequencies to mini-
mize the total cost with a metaheuristic algorithm.
4 - Inventory Pooling And Transshipment Under Correlated
Fat-Tail Demands
Zhen Liu, Numerix LLC, 1237 Bristol Ln, Buffalo Grove, IL, 60089,
United States,
We study the classic inventory pooling problem by Eppen (1979) under a special
class of multivariate fat-tail distribution: Normal Inverse Gaussian (NIG) demands
to better fit real-world demand data. We obtain the optimal inventory level in a
closed form by employing standardized NIG density function, and express the
optimal expected costs in terms of unit NIG loss function. Rather than
independent and identically distributed demands, our results complement
Bimpikis and Markakis (2015) by considering correlated demands. We further
discuss the transshipment problem of Dong and Rudi (2004) under NIG demands.
5 - Cyclic Vs. Static Inventory Policy Assumptions When Optimizing
Case-pack Sizes In Grocery Retailing
Heinrich Kuhn, Catholic University of Eichstaett-Ingolstadt, Auf
der Schanz 49, Ingolstadt, 85049, Germany, ,
Thomas Wensing, Michael Sternbeck
We analyze two possible approaches quantifying the optimal case-pack (CP) size
in retail distribution systems. One approach assumes a cyclic inventory policy
taking into account the weekly seasonality of product demands. The other
approach assumes an equal demand distribution on each day of the week, i.e.,
assuming a stationary inventory policy. The general assumptions of a periodic
review reorder inventory policy, i.e., (r, s, nq) policy, are assumed. We conduct
several experiments analyzing the question under what circumstances both
approaches achieve equal results.
MC56
Music Row 4- Omni
IT and Services
Sponsored: EBusiness
Sponsored Session
Chair: Atanu Lahiri, University of Texas at Dallas, Richardson, TX,
United States,
1 - Data-driven Optimization: Revenue Analytics For A Supply-side
Network Via Traffic-stream Mixing
Zhen Sun, University of Texas-Dallas, 800 west campbell rd,
Richardson, TX, United States, , Milind
Dawande, Ganesh Janakiraman, Vijay S Mookerjee
This study develops a data-driven approach to solve constrained optimization
problems in which the decision maker does not have an analytic form for the
objective function, but knows what decision variables affect the function. Our
approach comes with a worst-case performance guarantee that improves with the
characteristics (size, pervasiveness) of the available data. We apply our technique
to a traffic-stream mixing problem encountered by a supply-side internet
advertising network that wishes to optimize the click revenue earned from ads.
2 - Cardinality Bundles With Complex Costs
Jianqing Wu, Foster School of Business, University of Washington,
Seattle, WA, United States, , Mohit Tawarmalani,
Karthik Kannan
This paper studies pricing of cardinality bundles (CB) when bundling involves
complex costs. When implementing CB, a firm set prices on the sizes of bundle
and lets consumers choose specific products based on bundle sizes. The basic
model of CB is analyzed in Wu et al. (2016). In this paper, we first extend the
existing CB model to allow fixed costs in adding additional bundles. We show that
CB problem with fixed costs can be solved as a shortest-path problem. We then
extend the CB model in another way to solve CB problem with submodular cost
structure. Such analysis is especially useful when there exists economies of scale
in production.
4 - An Economic Analysis Of The Impact Of Recommender Systems
On Product Search
Abhijeet Ghoshal, University of Louisville,
Vijay M. Mookerjee, Sumit Sarkar
We perform an economic analysis of firms providing recommendation services
considering the influence of recommendations on the search process of
customers. We determine the price and recommendation system effectiveness
equilibrium and analyze how the equilibrium shifts when costs of
recommendations change.
MC57
Music Row 5- Omni
Experiments in Supply Chain Management
Sponsored: Behavioral Operations Management
Sponsored Session
Chair: Andrew M. Davis, Cornell University, Ithaca, NY, United States,
1 - Experimental Evidence On Post-choice Forecasting Bias:
Do Optimizers Know They’re Cursed?
Jordan Tong, University of Wisconsin,
Choose the best alternative and predict its outcome. We find that people tend to
be too optimistic in such post-choice forecasting tasks. This tendency has
important implications. For example, it suggests that managers who pick products
to include in assortments tend to forecast higher demand and order more
inventory than managers who don’t pick. We develop a behavioral model to help
explain the phenomenon and provide supporting experimental evidence in three
settings: guessing the number of pennies in jars, forecasting and making
inventory decisions for products, and estimating sales prices of houses. Finally, we
study factors that exacerbate the bias and possible ways to mitigate it.
2 - Firm Objectives And Managers’ Pricing Decisions:
Theory And Experiments
Rashmi Sharma, Penn State University, ,
Saurabh Bansal, Elena Katok
In this paper we develop a model to determine what compensation plan a firm
should offer to managers who make pricing decisions under price-responsive,
uncertain demand. We show that the structure of the compensation plan depends
on the firm’s objective function. We then report the results of a behavioral
experiment to test the model’s predictions.
3 - Multidimensional Bargaining In Supply Chains:
An Experimental Study
Kyle Hyndman, University of Texas at Dallas,
, Andrew M. Davis
We experimentally investigate the impact of bargaining and the allocation of
inventory risk on the performance of a two-stage supply chain. We show that
when allowing the parties to bargain over all contract terms simultaneously,
observed supply chain efficiencies are at least 90%, which are considerably higher
than those seen in past studies. Second, contrary to the theoretical prediction, the
party incurring the inventory risk always earns a substantially lower profit than
the other party. Third, a win-win situation is created when all contract terms are
simultaneously negotiated.
4 - Group Identity To Manipulate Social Preferences In Sales And
Operations Planning
Felix Papier, Associate Professor, ESSEC Business School,
3 Av Bernhard Hirsch BP 50105, Cergy Pontoise Cedex, France,
, Ulrich Thonemann, Torsten Gully
We analyze a supply chain in which a demand planner provides demand forecasts
to a production planner. The production planner needs information about the
forecast accuracy. If the actual effort of the demand planner and the belief of the
production planner are not aligned, the supply chain performance suffers. We
develop a game theoretic model to show how social preferences affect the
alignment between the two supply chain actors. Using lab experiments we find
that some demand planners invest effort, and that production planners anticipate
this effort. We further show that group identity can increase social preferences,
which ultimately leads to higher supply chain profit.
MC57