2016 INFORMS Annual Meeting Program

TA49

INFORMS Nashville – 2016

TA50 212-MCC SpORts: Sports Analytics Education

3 - Advances In Social Media Analytics For Ultra-low Latency Fintech Applications Martin M Spollen, Queen’s University, Belfast, BT9 5JX, United Kingdom, m.spollen@qub.ac.uk This presentation will discuss some recent advances in adaptive machine learning designed to extract more reliable capital markets foresight from streaming social media firehose data. Such techniques must keep pace with the rapidly evolving human language used on social media and deliver outputs with a minimum of latency for real time applications for Hedge Funds and High Frequency Algorithmic Traders. 4 - Momentum In Social Media And Sale Performance After Automobile Recalls Yen-Yao Wang, Michigan State University, N204 Business College Complex, East Lansing, MI, United States, wangyen@broad.msu.edu, Tawei Wang, Roger Calantone Due to the unique nature of social media, many firms have turned their attentions to social media to manage their recall campaigns. However, the role of social media before and after recalls has not received a more detailed examination. The purpose of this paper is to (1) assess the impact of social media on customers’ attentions to US mid-size vehicle recalls, and (2) examine the role of momentum in social media before and after the recall process. We obtained all mid-size automobile recall events and supplemented with social media data on customers’ discussions on defected vehicles and firms’ recall process from around 1,000 different social media platforms from 2010 to 2015.

Sponsored: SpORts Sponsored Session

Chair: Keith A Willoughby, University of Saskatchewan, 25 Campus Drive, Saskatoon, SK, S7N 5A7, Canada, willoughby@edwards.usask.ca 1 - Bracketology: How Business Analytics Can Help You Fill Out Your Bracket Michael Magazine, University of Cincinnati, mike.magazine@uc.edu This course is open to advanced undergraduates/graduate students with at least one course in probability and statistics. It meets over three Saturdays -one each in February, March and April. The course covers research papers that both determine the probability that one team beats another and also how brackets should be formed and filled out. One class is devoted to student teams acting as the selection committee and justifying how they form brackets. Students and instructors (I co-teach this with Paul Bessire, an ex-student who is founder of Predictionmachine.com) compete in a bracket challenge and prizes awarded to the best performers. The last class has included visitors, like Joe Lunardi of ESPN. 2 - When Is It Ok Not To Score? Teaching Decision Analysis With The Sport Of Curling Keith A Willoughby, University of Saskatchewan, Saskatoon, SK, Canada, willoughby@edwards.usask.ca, Kent J. Kostuk The object of sports is to outscore your opponent. Curling is a winter team sport popular in Canada, Europe, the northern United States and the Pacific Rim. In the sport of curling, teams may encounter a crucial decision in the latter stages of the game; namely, should they score a point (thereby providing last-shot advantage to the opposition in subsequent stages of the game) or deliberately fail to score a point (thus retaining last-shot opportunity in the next part of the match)? We develop a model for this particular scenario that can be used to teach introductory decision analysis. 3 - A Playbook For Teaching Sports Analytics To Undergraduate Business And MBA Students Scott Nestler, University of Notre Dame, snestler@nd.edu Last year, the presenter taught a half-semester length course (2 credits for MBA students, 1.5 credits for undergraduates) in Sports Analytics for the first time at the University of Notre Dame. Offensive plays — reaching out to faculty members at other schools who had taught a similar course yielded many great examples; allowing students freedom to use whatever tool or coding language they were comfortable with for the course project. Defensive plays — selecting Excel as a common language for class examples to account for difference in technical preparation (may revisit for next semester with a more pro-style scheme that incorporates R); using a known but somewhat dated text (Winston’s “Mathletics”). Please come listen and your experiences in teaching quantitative techniques using a subject matter that students are truly excited about. Lifting up Populations Sponsored: Public Sector OR Sponsored Session Chair: Feyza Guliz Sahinyazan, McGill University, Desautels Faculty of Management, Montreal, QC, HA 1G5, Canada, feyza.sahinyazan@mail.mcgill.ca 1 - Resilience-based Post-disaster Recovery Strategies For Community Road-bridge Networks Weili Zhang, University of Oklahoma, 202 W. Boyd St., Room 116, Norman, OK, 73019, United States, weili.zhang-1@ou.edu, Naiyu Wang, Charles Nicholson This paper presents a novel resilience-based framework to optimize the scheduling of the post-disaster recovery actions for community road-bridge transportation networks. Two metrics are proposed for measuring rapidity and efficiency of the network recovery: the TRT is the time required for the network to be restored to its pre-hazard functionality level, while the SRT is a metric defined for the first time in this study to capture the characteristics of the recovery trajectory that relate to the efficiency of those restoration strategies considered. Based on this two-dimensional metric, we propose a restoration scheduling method for optimal post-disaster recovery planning. TA51 213-MCC

TA49 211-MCC Classroom Activities Sponsored: Education (INFORMED) Sponsored Session

Chair: Vincent Hargaden, Assistant Professor, University College Dublin, 209 Engineering & Materials Science Centre,, Belfield, 00000, Ireland, vincent.hargaden@ucd.ie 1 - An Interactive Spreadsheet Based Game For Teaching Design Of

Experiments And Response Surface Methodology Anthony Bonifonte, Georgia Institute of Technology, ABonifon@gatech.edu

Experimentation is a key feature of many scientific and engineering disciplines. This presentation describes an interactive spreadsheet based game implemented in a quality control course. The game is designed to simulate an industrial or laboratory experimentation process and develops skills in design of experiments, response surface methodology, optimization, and statistical analysis. The game is appropriate for the undergraduate or masters level and relevant for any course that teaches experimentation. 2 - Using Jupyter Notebook In The Operations Research Classroom Nelson A Uhan, United States Naval Academy, uhan@usna.edu Jupyter Notebook is an interactive computational environment that allows you to create documents that contain live code, text, equations, and visualizations. As a result, Jupyter Notebook can be a very useful teaching and learning tool for classes with a considerable emphasis on programming and computation. In this talk, I will share my experience with using Jupyter Notebook in undergraduate operations research classes, and discuss some of my plans for using it in the future. 3 - Analysis And Design Of Discrete Material Flow Systems: A Virtual Industrial Engineering Systems Pilot Laboratory United States, dima.nazzal@isye.gatech.edu, Leon McGinnis, Timothy Sprock, George Thiers In this project we redesigned a core undergraduate course that focuses on the analysis and design of discrete material flow systems. We partnered with MathWorks to use Matlab and created a virtual Industrial Engineering systems lab; a suite of computational components that enable students to “experiment” not just with the kinds of analytic models we routinely teach, but also with computational models of versions of the systems they represent where the simplifying assumptions are relaxed. This talk will illustrate samples of the computational tools we developed and how they were integrated into the course and utilized to enhance students’ understanding of the key concepts covered in this course. 4 - Teaching Earned Value Analysis Using A Classroom-based Dice Game Vincent Hargaden, Assistant Professor, University College Dublin, 209 Engineering & Materials Science Centre, Belfield, Ireland, vincent.hargaden@ucd.ie, Virpi Turkulainen We describe the use and evaluation of a classroom based dice game to teach the concept of Earned Value Analysis. A summary of the game and teaching materials will be outlined. We describe how the perceived effectiveness of the game as a teaching tool was measured among different cohorts of students.

250