|
School of Operations Research and Industrial
Engineering
College of Engineering, Cornell University |
|
|
|
ORIE 2002 Newsletter
ORIE 2002 Newsletter
Contents
Spring Semester Highlights New Faculty Member: The academic year has been lively and a bit exhausting. We conducted multiple faculty searches, which always strain people's time and intestinal fortitude. The searches yielded one success and we have landed a gem from Princeton, Huseyin Topaloglu, who starts his Cornell career in August. Huseyin is an exciting catch for us. He is very broad and flexible, has industrial experience, great enthusiasm and is just what is needed to strengthen our applied operations research and systems engineering activities. The other searches were unsuccessful for a variety of reasons: two-body issues, location, financial competition from competitors. We will continue efforts to hire the right people in the coming academic year. David Ruppert Named Andrew Schultz, Jr. Professor of Engineering: Professor David Ruppert, ORIE, has been appointed to the Andrew Schultz, Jr. Professorship of Engineering. The Andrew Schultz, Jr. Professorship in Engineering was established in 1980. Andrew Schultz came to Cornell in 1932 as a student in Mechanical Engineering. At Cornell, Andy received the BS in Administrative Engineering in 1936 and the Ph.D. in 1941. He served as Assistant, then Full Professor, in CornellĘs Industrial Engineering and Administration Department and became department head in 1951. Andy served as fifth dean of Cornell's College of Engineering from 1963-1972 and returned as acting dean in 1978. The endowed chair is designated for a "leader in the application of both industrial and engineering operations research approaches to problems arising in manufacturing and service organizations." Dave Ruppert received his BA from Cornell in 1970 and Ph.D. from Michigan State in 1977. He was Assistant and then Associate Professor in the Department of Statistics, University of North Carolina, Chapel Hill from 1977-1987 and returned to Cornell as Professor in the School of Operations Research and Industrial Engineering in 1987. He is known for his contributions to generalized and non-linear regression, robustness theory, data transformation and weighting, non-parametric function estimation and measurement error modeling. His research exemplifies good taste and sits at the boundary between deep theory and genuine applicability. He is known for practical statistical solutions to challenging practical problems as well as the depth of his fundamental contributions through deep mathematical analysis, which proves the value of a suggested methodology. He has influenced standard practice and education by the publication of three books with a fourth in the contract stage. He has guided 17 Ph.D. students, many of whom have gone on to professional prominence and influence. Both leading statistical societies, The Institute of Mathematical Statistics and the American Statistical Association, have awarded Professor Ruppert Fellow status, their highest honor. He has been associate editor of all the leading professional journals (Biometrics ('99 -present), Journal of the American Statistical Association ('88-'96), Annals of Statistics ('89-'91), American Statistician ('83-'87). He was also a two-term editor-in-chief of the Lecture Notes and Monograph Series of the Institute of Mathematical Statistics. He is also in demand to give short courses and has given three in the last three years. He is a frequent participant in NIH and NSF panels. Twenty-five years ago Andy Schultz pointed out that Engineering was filled with people teaching "rules-of-thumb" engineering without enough quantitative justification for the uses of these ideas. He thought engineering would better progress if people understood how to properly extract information from data. Andy, and many Cornell alumni, felt one of the cornerstones of the School of Operations Research and Industrial Engineering was proper understanding of the importance of acquiring and scientifically analyzing data. Professor Ruppert's stature, accomplishments, breadth, Cornell commitment, professional vision and judgment will make him a distinguished holder of the Schultz chair who will do justice to Andy's broad and clear vision. Financial Engineering Presentation to the College Council: Jeremy Staum gave a presentation to the Engineering College Council on April 10, updating the College's advisory body on developments and plans in Financial Engineering. A follow-up April 25 meeting with Jay Carter (BS '71, M. Eng. '72), Senior VP, Lucent and a College Council member, was useful feedback as suggestions were made for how the program could be broadened in the future. The School's Financial Engineering Program has been stabilized under the directorship of Philip Protter and he, Professor Staum, and Associate Director for M. Eng. Studies Mark Eisner, have been thinking of ways to broaden and improve the program. The launching of a new Financial Engineering web page has been accomplished (access at http://www.orie.cornell.edu:4000/) and Professor Protter has been the mover behind the establishment of a data repository to provide project material and test bed environments for new procedures. Local press coverage of M. Eng. project: On April 30, The Ithaca Journal highlighted a M. Eng. project to develop a web site trip planner for the local bus system TCAT (Tompkins County Area Transportation). Dwight Mengel, TCAT's service development manager commented that the user-friendly tool designed by the students "is evidence of some of the tremendous assistance that such a team of capable students could give to any community organization." The article appeared at http://www.theithacajournal.com and was cited during Program Review. Such School projects have also been discussed in connection with Cornell's efforts to reassess outreach and economic development in New York State and to create a modern definition of a land grant institution. President Hunter Rawlings stimulated such a reassessment prior to his announcing retirement from the Cornell presidency effective June 2003.
and "Functional Data Analysis" and wrote several technical reports with Cornell faculty. (Technical reports are available on our web page www.orie.cornell.edu.)
Jack Muckstadt was voted "Outstanding Professor of the Year 2002" by our chapter of the American Institute of Industrial Engineers. The Master of Engineering Students voted their Outstanding Teaching Award 2002 to (you guessed it) Jack Muckstadt. Congratulations to Jack. Where does Jack store all the plaques? Ganesh Janakiraman and Joerg Rothenbuehler were voted Outstanding Teaching Assistants by the Master of Engineering and undergraduate students, respectively. Congratulations to all for well deserved recognition. In the coming fall semester Jack Muckstadt and Robin Roundy will be honored by the College as Distinguished Teaching Award winners.
Emeritus Faculty: N.U. Prabhu. Leslie Trotter contributes the following.
Despite his official retirement nearly a decade ago, Emeritus Professor N. U. ("Uma") Prabhu's commitment to the School's academic activities has not diminished. He still teaches regularly, offering for this coming fall semester a Senior/M.Eng. level edition of his signature course, OR561: Queueing Theory. Uma also recently directed an M. Eng. project group designing software for capacity planning for processing credit card requests, and he is currently overseeing the academic progress of a group of junior ORIE majors, shouldering part of our faculty's disproportionate share of advising responsibility caused by sharply increasing numbers of undergraduate majors in the School.
This year marks two significant anniversaries in Uma's remarkable career. First, the initial steps of his academic journey began 50 years ago with an appointment in the Department of Statistics at Karnatak University in India. Then, as now, Uma felt that direct interaction with students was an indispensable component of scholarship. This dedication to teaching has sustained his efforts throughout these 50 years, never being more clearly evidenced than by his present contributions. In his own words, "I have always felt a very personal commitment to teaching &cdots; much more than a simple job-related obligation which could be shelved at age 65." The present academic year is also the 35th anniversary of Uma's promotion to Professor here at Cornell. He arrived at Cornell in 1965, during the height of a period of far-reaching changes in the College of Engineering -- this was the year in which the Department of IE&OR was established and its undergraduate four-year B.S. degree initiated, as well as the year which saw the creation of the Department of Computer Science. These changes had been set in motion by the (then) Dean of the College of Engineering, Andy Schultz, as part of Cornell's role in what he would later characterize as the quiet revolution from scientific management to operations research (see Engineering: Cornell Quarterly, Vol. 4, No. 4, Winter 1970). For our academic unit this was a time of expansion and modernization, ultimately to be reflected a decade later in 1975 by the official inauguration of the School of ORIE. Uma's appointment was representative of the strong commitment in this direction shared by the College and the School.
Sarah Hattery (ORIE, '03), under the guidance of David Ruppert (ORIE), submitted a successful summer funding proposal to the College. The Learning Initiatives for Future Engineers (LIFE) Program approved the proposal for the General Electric Faculty for the Future Summer Research Program. Kasey Kudamik (ORIE, '02) received an NSF Graduate Fellowship to pursue a Ph.D. in Environmental Engineering and also was awarded the Byron W. Saunders Award as our outstanding senior with the highest grade point average. Two prizes awarded to ORIE graduates continuing on to the M.Eng. program include the Lynn E. Bussey Scholarship given to Hsiang-Yi Lin and the Allan H. Mogensen Award given to both Ernest Fung and Justin Maynard.
Ciprian Crainiceanu (Ph.D. candidate, Statistics, working with David Ruppert) won "Best overall student presentation" for the paper Ciprian Crainiceanu, David Ruppert, Jery Stedinger, Christopher Behr, "Bayes Analysis of Pathogen Risk in Natural Water Supplies" at the conference "Applications in Statistics: What's Hot/ What's Next VII, Statistics and Public Health," Albany, March 23, 2002. The Conference was organized by the Albany Chapter of ASA, School of Public Health at the University of Albany. The research is interdisciplinary and is sponsored by an EPA grant to Stedinger and Ruppert. The work provides a statistical basis for assessing health risks due to pathogens in drinking water supplies.
Sherri K. Stuewer (ORIE, '73) presented the 2001-2002 John R. Bangs Jr. Memorial Lecture on April 10 on the topic "Optimization in the Oil Business from Top to Bottom." Stuewer is strategic planning manager at ExxonMobil Corporation in Irving, Texas. She is also chair of the Engineering College Council. The Bangs lectureship was established in 1990 to honor the Cornell professor who worked to establish the field of administrative engineering. One of the goals of the lectureship is to provide periodic contacts for students, faculty and staff with experts in the field of production operations. Network upgrades: The network in our computing labs has been upgraded from 10 to 100MB. Two Windows servers have been added and operating systems updated. The fiber optic link between Rhodes Hall and the Upson Hall M. Eng. computer lab was updated from 10MB to 100 MB, 320GB of additional user storage was made available for user home directories, and we purchased a 40GB/80GB Quantum tape back unit to improve system backup capabilities. Two new 1.4 GHz laptops for student presentations and classroom use have been purchased to replace the creaky ones that received constant use.
Mark Eisner, Associate Director for the Master of Engineering Program, reports the following. A large Master of Engineering class participated in a broad array of projects this past year. Clients ranged from the local bus system (see the second page of this letter) and the Geneva Experiment Station to financial consulting, software, supply chain, and pharmaceutical companies. Many alums have found it rewarding to get involved in the sponsorship of M. Eng. projects, directly or through their employers. It's a great way to get fresh insights, under faculty guidance, into a wide range of business problems. We have recently prepared a brief description of the process by which projects are sponsored. Contact Kathy King, kathy@orie.cornell.edu, for an electronic or paper copy of this document which can also be accessed at our web site http://www.orie.cornell.edu/meng/sponsoringmengproject.html. Good projects are always needed for M. Eng. teams. The best projects are those which are important enough to the client that there is a willingness to provide management time to work with the students and data for them to use, but not so important that the rare instance of poor performance puts the client at business risk. Clients are expected to pick up out-of-pocket project costs such as travel, telephone, reproduction and any acquired software. If you have an interesting project in mind, contact Mark Eisner, Associate Director for the M. Eng. Program (Eisner@orie.cornell.edu) or phone 607-254-3317. Some examples of this year's projects are given next. In another project with a client close at hand, Dan Dalthorp advised the team consisting of Mary Kiang and Patricia Moll that developed a method that might reduce the amount of pesticides used to control grubs on golf course fairways. This project was done for Cornell's Agricultural Experiment Station in Geneva, at the north end of Seneca Lake. Rather than indiscriminant use of pesticides for prophylactic reasons, the team developed a way to help identify those fairways that really need treatment to fight grubs. The method is based on sampling to predict how critical the grub problem will be in a particular year and takes into account the economic tradeoff between sampling and treatment as well as the risk tolerance of the individual golf course manager. In the method, samples are analyzed using an add-in to the Matlab statistical software package. Customer segmentation has become a crucial technique in marketing many products since it enables firms to tailor their marketing strategies to match the attributes of different customer segments. For a given prospective customer, it is possible to determine membership in a particular market segment by obtaining responses to a detailed questionnaire, but this approach is prohibitively expensive. The client, a strategic marketing firm, determined segment membership from questionnaire responses for a set of individuals. The client also had detailed information, collected in the ordinary course of business, about their behavior as investors in a mutual fund company that the client works with. The team consisting of Mike Jeun, Allen Liou, Theo Rusli and Stephen Terry, under the guidance of Wenxin Jiang and Mark Eisner, found out how well several different methods predicted membership based on the behavioral data instead of the questionnaire data. The project determined that a neural network approach not previously used by the client showed the most promise in this task. Authoria, Inc. provides software to the human resources function of major corporations that they use to manage self-service web sites about employee benefits. With mergers, acquisitions and other changes, employees for a single company may have different benefit coverages. Making sure each employee sees precisely the right information about their own benefits is a key value of this software, but only if the software itself produces accurate information free of errors ranging from typographical to logical. Quality Assurance at Authoria is resource intensive, with a varying level of activity depending on the release schedule of the software. The project team consisting of Amjad Alinani, Roman Cavelle, Micolas Malfroy-Camine and Tu Tran under the guidance of Mark Eisner, developed ways in which Authoria can use past experience to predict the level of QA effort required and to help allocate work to employees. They helped modify the existing error tracking program to make defect data collection easier and to pass the data to the program that forecasts defect rates and testing times. Project team members, who used ORIE techniques for a software developer, were in the new Information Technology M. Eng. option. As ORIE alums know, "making intelligent strategic, tactical and operational decisions in the presence of uncertainty and variability is paramount to success in today's business world." This quote from the Executive Summary sets the stage for describing an investigation of the impact of uncertainty on design and operational decisions for supply-chains. Often, strategic planning deals with uncertainty by running deterministic models through a variety of scenarios. Working with ORIE Ph.D. alumnus Dwight Collins of Aspen Technology and Jack Muckstadt, the project team of Pat Chen, Ryan Titchett, Robert Lien and Peer Mumme conducted simulation experiments to investigate the implications for this approach of forecasts that fail to accurately predict demand rate, of decisions that do not take the variability of operating parameters into account, and of supply-chains that incorporate the collaboration of multiple companies. With respect to these sources of variability, the team concluded that the value of strategic planning models could be greatly enhanced if they incorporate uncertainty through simulations. Clinical trials of prospective new products are a key step bringing pharmaceutical products to market. Trials are expensive and time-consuming, and require resources. There are tradeoffs between staffing levels and organizational performance. Organizational units that determine the budget for clinical trials must deal with these tradeoffs. A project for Pfizer Pharmaceuticals developed a simulation tool to estimate completion times, analyze workflow, evaluate resource levels and spotlight opportunities for process improvement in the budgeting process. Shane Henderson advised project team members Evan Glisson, Jong Chul (Andrew) Kim, Patrick Ryan and Chia-Feng (Daniel) Lau. Information technology has made it easy for many vendors to bid on raw material orders in manufacturing, using an auction format. Robin Roundy supervised a project based on some of his research to analyze a particular form of auction used in "B2B" transactions, known as a Vickrey auction. Vickrey auctions are designed to provide an incentive for bidders to bid their true cost, through the use of sealed bids with an interesting wrinkle ū the buyer picks the low bidder but pays that bidder the price bid by the second-lowest bidder. Game theory can be used to show that the Nash equilibrium is for bidders to bid their true cost. However the analysis gets complicated if multiple items are being auctioned in a single round. Moreover these auctions can entail "side" payments by the seller to the buyer for the right to participate, and price levels ("reserve prices") beyond which the buyer is unwilling to go. The team, Fernando Abadia, Tae Hoon Kim, David Kull and Leonardo Widjaja, developed a simulation to investigate the effects of side payments and reserve prices and to recommend the optimal values for both.
In addition to these and other projects, teams in the Financial Engineering option carried out projects for Lehman Brothers and Morgan Stanley, and ORIE M. Eng. students joined students from other M. Eng. programs on teams in the Systems Engineering and Manufacturing Options.
Alper Yildirim completed his dissertation "An Interior-Point Perspective on Sensitivity Analysis in Linear Programming and Semidefinite Programming"
under the supervision of Michael Todd, Alper is now on the faculty as an Assistant Professor in the Applied Mathematics department of the SUNY at Stony Brook.
Barbara Gonzalez dissertation entitled "Performance of a Leaky Bucket System with Long-range Dependent Input Traffic" was done under the supervision of Gennady Samorodnitsky. Her work studies the effect of admission protocols in a communication network with stochastic service times, in which the service times are characterized by heavy tails and inputs have long memory. Barbara has accepted a position as an Assistant Professorship of Statistics at the University of Louisiana at Lafayette.
Funding Priorities for the Immediate Future As outlined in our Strategic Plan, our traditional instructional and research programs are very popular with students. The number of majors per tenured faculty member far exceeds any other unit in the College of Engineering to the point where large class size is one of our important concerns. Despite this success and popularity, our School considers it vital to expand our focus both for the sake of our students and for the continuing vitality of our subject of Operations Research. This recognizes a changing world in which service and information become paramount. To prepare for new opportunities provided by these changes, our students need a broader offering and set of choices than is provided with our traditional orientation. Some of the areas that we have been developing are financial engineering; information systems engineering and economics applied to networks, supply chains and inventory analysis; data mining applied to business and financial services. We need to continue to develop programs in these new areas and we seek support designated for one of our new programs. Support could be targeted to benefit (a) financial engineering. (b) information technology and economics. (c) data mining and its applications and other modern statistical methodologies. (d) infrastructure needs such as computing lab renovation, software licenses, computing renewal, physical plant maintenance. Support for a developing program could be used by the Director and program leaders in a number of ways such as: (i) faculty support. We need to expand our faculty both to gain additional expertise and to cope with exploding enrollments. (The number of new affiliates per year to our major used to range between 90-110. Now it has shot up to 140-165 per year.) We are currently seeking ways of paying for additional faculty beyond the core faculty supported by the college. (ii) graduate student support for students engaged in a research program attached to one of the emerging areas. (iii) course development support. (For example, faculty have attended data mining workshops, been given time off for developing courses such as "Quantitative and OR Methods Models in Finance.")
|
 People
 Courses
Employment Opportunities
Academic Programs
Financial Engineering
Technical Reports
Seminars
Strategic Plan
Newsletters
ORIE in the News
ORIE only: policies, services
Useful Links
Request Information
|