Informs 2015

400 / 552

INFORMS Philadelphia – 2015
398
WA74
74-Room 204A, CC
Decision Analysis Approaches and Predictive
Modeling to Managing Uncertainty in Manufacturing
and Service Systems Design & Operations
Sponsor: Quality, Statistics and Reliability
Sponsored Session
Chair: Zhenyu Kong, Associate Professor, Virginia Tech University, 123
Durham Hall (MC 0118) 1145 Perry Str, Blacksburg, VA, 24061, United
States of America,
1 - Self-Organizing Network for Variable Clustering and
Predictive Analytics
Gang Liu, University of South Florida, 4202 E. Fowler Ave.
ENB118, Tampa, FL, United States of America, ,
Hui Yang
This paper presents a new methodology of self-organizing network for variable
clustering and predictive modeling. Specifically, nonlinear interdependence
structures are characterized and measured among variables. Further, we
embedded these variables as nodes in a complex network. Nonlinear-coupling
forces move these nodes to derive a self-organizing topology of network. As a
result, sub-network communities delineates the cluster structure of variables.
2 - A State Space Model for Multivariate Poisson Count Series
Nan Chen, National University of Singapore,
Chen Zhang
This paper proposes a state space model to model and monitor multivariate count
series. The model is built on multivariate log-normal mixture of independent
Poisson distribution and allows for serial dependence by considering the Poisson
mean vector as a latent process driven by a state space model. Based on this
model, we furthermore propose a statistical process control scheme for
multivariate count series. Finally we present a case study in an electrical system.
3 - Generalized Wavelet Shrinkage of Raman Spectra for Quality
Control of Continuous Nanomanufacturing
Xiaowei Yue, Georgia Institute of Technology, 755 Ferst Drive
NW, Atlanta, GA, 30332, United States of America,
, J.g. Park, Jianjun Shi, Hao Yan, Kan
Wang, Chuck Zhang, Zhiyong Liang
Raman spectroscopy is an attractive in-line quality characterization and
quantification tool for continuous nanomanufacturing.A signal analysis model
with signal-dependent noises for Raman spectra is developed and validated.And a
novel generalized wavelet shrinkage approach is introduced to remove noise in all
wavelet coefficients by applying individual adaptive wavelet thresholds. The
effectiveness of this method is demonstrated using both simulation and
experimental case studies.
4 - A Novel Bayesian Learning Algorithm for Sparse Estimation under
Mixed Gaussian-Laplacian Noise
Babak Barazandeh, Research Assistant, Virginia Tech University,
Room 211 Graduate life center- 155 Otey, Blacksburg, VA, 24061,
United States of America, Zhenyu Kong,
Kaveh Bastani
Sparse estimation problem is very challenging due to the ill-posedness of the
problem. Typically the sparsity assumption provides some viable algorithms to
approximate the solution. However, most of these algorithms minimize the l-2
norm error due the Gaussian noise assumption, which is sensitive to outliers
caused by Laplacian noise. We propose a novel Bayesian learning algorithm with
a parameterized prior encouraging sparsity under the mixture of Gaussian and
Laplacian noise.
WA75
75-Room 204B, CC
Supply Chain Management IX
Contributed Session
Chair: Wan Libing, Huazhong University of Science & Technology, 1037
Luoyu Road, Wuhan, P.R. China, Wuhan, China,
1 - A Mathematical Model of Vehicle Routing Problem for Third-Party
Logistics Service Providers
Birasnav M, New York Institute of Technology, Broadway,
61st Street, Manhattan, NY, 10023, United States of America,
Third-party logistics service providers play a major role in the distribution of
matchboxes from southern towns to northern towns of India. These service
providers operate a facility called cross-docking that receives bundles of
matchboxes from various suppliers located around the facility and consolidates
and dispatches these bundles to customers.
2 - Multi Region Input Output Model for Scope-based Carbon
Footprint Accounting
Murat Kucukvar, Assistant Professor, Istanbul Sehir University,
Altunizade, No 27, Istanbul, 34662, Turkey,
Vural Aksakalli
This research aims to advance the body of knowledge on carbon footprint analysis
of the manufacturing sectors with a holistic approach combining the scope-based
carbon footprint accounting standards with a time series multi-region input-
output (MRIO) analysis framework. To achieve this goal, a global scope-based
carbon footprint analysis of the Turkish manufacturing sectors has been
conducted as a case study.
3 - Optimal Remanufacturing Strategies in Name-your-own-price
Auctions with Limited Capacity
Qifei Wang, Huazhong University of Science and Technology,
Luoyu Road, 1037 Building, Wuhan, Hubei, Wuhan, China,
, Chengxiu Gao, Hong Yan, Stuart X. Zhu,
Jianbin Li
We study optimal remanufacturing strategies in a closed-loop supply chain. The
returns are collected under a name-your-own-price (NYOP) bidding mechanism.
We find the manufacturer’s optimal strategies mainly depend on bidding cost, cost
saving, production capacity, and market scale. We also compare this mechanism
with traditional list price mechanism and find the conditions under which the
manufacturer prefers the NYOP mechanism. Some managerial insights are
obtained through numerical studies.
4 - Optimal Price and Sales Effort Level in a Dual-channel Supply
Chain in Case of Bidirectional Free Riding
Mengping Zhu, Huazhong University of Science and Technology,
Luoyu Road, 1037 Building, Wuhan, Hubei, Wuhan, China,
, Jianbin Li, Bin Dai
We study the optimal policy in case of bidirectional free riding and price
competition. Results show that: in the decentralized supply chain, retailers’
optimal sales effort depends on wholesale price, effort cost and the free riding
coefficient: when the effort cost is too high or too low, the manufacturer will offer
the same wholesale price for both retailers, otherwise, the manufacturer engages
in wholesale price discrimination.
5 - The Competition Strategy Analysis in a Mixed Duopoly Market
Wan Libing, Huazhong University of Science & Technology, 1037
Luoyu Road, Wuhan, P.R. China, Wuhan, China,
So far, China’s express industry has formed a mixed oligopoly market in which
state-owned enterprise and private enterprises compete together. This thesis
studies duopoly express service providers’ price and capacity based on
mathematical models. We compare social welfare between the two competition
strategies of state-owned company. The result shows that state-owned company
should either maximum social welfare or maximum profit as its objective which is
depended on technological conditions.
WA76
76-Room 204C, CC
Simulation I
Contributed Session
Chair: Pratik Misra, Air Products and Chemicals, Inc., 7201 Hamilton
Boulevard, Allentown, PA, 18195, United States of America,
1 - Optimal Design Models for 3d Automated Parking Systems
Guangmei Wu, Doctoral Student, Huazhong University of Science
&Technology, 1037 Luoyu Road, Wuhan, 430070, China,
Xianhao Xu, Yeming Gong
In this paper, we consider a newly designed for automatic parking garage system.
We mainly use queuing network models to estimate the system expected travel
time when we retrieve the cars. Our research objective is to analyze the system
performance and optimally design for automatic parking garage system.
2 - Efficient Sampling of the Brown-resnick Process
Zhipeng Liu, PhD Candidate, Columbia University, 424 West
119th Street, 6B, New York, NY, 10027, United States of America,
, Jose Blanchet
We propose a perfect simulation algorithm for the Brown-Resnick process. This
algorithm has an optimal computational cost in the sense that it has nearly linear
complexity with simulation of the basic Gaussian Random field. We mainly use
importance sampling and acceptance-rejection procedure to construct this
algorithm.
WA74