2015 Informs Annual Meeting

WC54

INFORMS Philadelphia – 2015

WC54 54-Room 108A, CC Service Science I Contributed Session

2 - Combining Forecast Quantiles – A Numerical Investigation Chen (mavis) Wang, Assistant Professor, Tsinghua University, Shunde Bldg. S609, Beijing, 100084, China, chenwang@tsinghua.edu.cn, Shu Huang, Vicki Bier We review statistical methods for combining forecast quantiles from multiple experts, including equal weighting aggregation (by finding the average quantile, average probability, median quantile, or median probability), and performance- based weighting aggregation. We also propose a Bayesian quantile regression model to estimate location and precision biases and a more flexible Bayesian nonparametric model. We compare them using both simulation and a large dataset on expert opinion by Cooke. 3 - Approximate Representation for Time Series and its Application to Efficient State Estimation Jianjun Lu, Associate Professor, China Agricultural University, No. 17, Qinghuadong Road, Haidian Distri, Beijing, 100083, China, ljjun@cau.edu.cn We propose an approximate representation of multivariate time series by using the representative time series called latent time series based on covariance structure analysis where correlation among observed time series is utilized. The dynamic factor analysis for multivariate time series is extended. To avoid the whole estimation of state variable to each nonlinear system, we apply Particle Filters only for latent time series, and for another observed time series we use these estimated states. 4 - Forecasting Trends of Immigration to United States of America Yudhi Ahuja, Associate Professor, San Jose State University, This paper deals with immigration statistics to United States of America from all over the globe. The future trends of immigrants have been estimated and their distributions over various States worked out. The paper concludes with implications of immigration on the economy, employment, education and social benefits. 5 - Integrated Fuzzy Approach for Analyzing Risk Analysis in Product Recovery System Dr Jitender Madaan, Professor, Dept Of Management Studies, Indian instititute of Technology Delhi, Hauz Khas, New Delhi, 110016, India, jmadaaniitd@gmail.com, Divya Chaoudhry Paper proposes proposes a novel methodology based on fuzzy set theory and evidential reasoning algorithm for quantifying the risks to capture the uncertainties of recovery system operations. It has practical implications for the organizations involved in product recovery to guide strategy formulation for the pro-active mitigation of risks. Moreover, paper would provide insights to the managers for enhancing the robustness of recovery systems along with better management of disruptions. One Washington Square, San Jose, CA, 95192, United States of America, yudhi.ahuja@sjsu.edu Chair: Felix Papier, Associate Professor, ESSEC Business School, Avenue Bernard Hirsch, Cergy, 95021, France, papier@essec.edu 1 - Managing Product Variety through Developing Vanilla Boxes using Hierarchical Clustering Pooya Daie, Concordia University, 1455 De Maisonneuve Blvd. W., Montreal, QC, H3G 1M8, Canada, Pooyadaie@gmail.com, Simon Li This paper focuses on implementing mass customization through development of semi-finished vanilla boxes to reduce supply chain cost.The challenge is that the possible number of vanilla boxes grows dramatically with increase in number of product variants.In solution,the basic information of product variety is captured in a matrix format,specifying the component requirements for each product variant.Then,hierarchical clustering is applied over the components with the considerations of demands 2 - Tailor Made: Make-to-order Decisions at the Bottom of the Apparel Supply Chain Suri Gurumurthi, Visiting Asst. Professor, HKUST, Clearwater Bay, Kowloon, HK, Hong Kong - PRC, imsuri@ust.hk I discuss make-to-order strategies considered by apparel component suppliers at the bottom of the supply chain. The prevalent models for make-to-order decisions focus on decision-makers at the upper tiers of supply chains. I discuss make-to- order decisions at lower tier decision-makers who are subject to extremes of demand variability, but that also compete with very slim margins. WC56 56-Room 109A, CC Manufacturing I Contributed Session

Chair: Oleg Pavlov, Associate Professor Of Economics And System Dynamics, WPI, SSPS, 100 Institute Rd., Worcester, MA, 01609, United States of America, opavlov@wpi.edu 1 - Demand Management and Optimal Workforce Scheduling in Professional Service Firms Vincent Hargaden, University College Dublin, Engineering & We analyse the workforce planning process in professional services firms from a demand management perspective. Taking a project network approach, we develop a multi-objective optimization model, which minimizes both the total direct costs of project staff and the maximum delay from the scheduled durations across all projects. The result is a better utilization of the firm’s workforce, while maintaining customer satisfaction. 2 - Strategies for Planning the ICT Convergence-based PSS Value Chain in Manufacturing Companies Hosun Rhim, Professor Of Logistics, Service, And Operations Management, Korea University Business School, Anam-dong, Seongbuk-gu, 136-701, Seoul, Korea, Republic of, hrhim@korea.ac.kr, Yong Yoon The ICT convergence-based Product Service Systems (PSS) is a migration strategy to be adopted for the manufacturers in the PSS business model planning stage. The structured process and the critical factors in that stage will be investigated. The analytic hierarchy process (AHP) has been implemented to organize and analyze a series of decisions. 3 - Exploiting Learning in Call Center Routing Decisions Tom Robbins, Associate Professor, East Carolina University, College of Business, 3212 Bate Building, Greenville, NC, 27858, United States of America, robbinst@ecu.edu We explore a call center environment where agents increase their productivity over time, but eventually quit. We consider a routing policy that attempts to exploit this situation and improve long-term call center performance. We examine policies where calls are routed to the most experienced agents when the call center is busy, to facilitate efficiency, and to the least experienced agent when Materials Science Centre, Belfield, Dublin 4, Ireland, vincent.hargaden@ucd.ie, Jennifer Ryan, Amir Azaron Oleg Pavlov, Associate Professor Of Economics And System Dynamics, WPI, SSPS, 100 Institute Rd., Worcester, MA, 01609, United States of America, opavlov@wpi.edu, Frank Hoy Service science is an emerging discipline rooted in system science. Applied to education, the service science methodology views universities and academic programs as service systems that go through life-cycles. We study entrepreneurship education programs that are gaining in popularity, yet are notoriously difficult to build up and sustain. We describe them as educational service systems, review different program deployment models and identify factors that lead to their success or failure. the call center is slow, to facilitate learning. 4 - Education as a Service System

WC55 55-Room 108B, CC Decision Analysis II Contributed Session

Chair: Yudhi Ahuja, Associate Professor, San Jose State University, One Washington Square, San Jose, CA, 95192, United States of America, yudhi.ahuja@sjsu.edu 1 - Social Norms and Identity Dependent Preferences Daphne Chang, School of Information, University of Michigan, 105 S. State Street, 3336 North Quad, Ann Arbor, United States of America, daphnec@umich.edu, Erin Krupka, Roy Chen In our paper, we test the impact of norms on social identity driven choice by using a 2 (identity prime) x 2 (frame) x 2 (choice or norms) experimental design to separately and directly elicit empirical measures of identity dependent norms for eleven different redistribution situations. We demonstrate that including identity dependent norms improves our ability to predict behavior. Further, we estimate a key structural parameter of the social identity modelóidentity dependent norm sensitivity.

444