Hooding Ceremony Honors New Ph.D. Graduates

Eight Ph.D. candidates from the United States, Canada, China and Sweden were recognized at ORIE's Graduation Ceremony in the Schwartz Auditorium on May 26.  Six are going on to careers in industry - at Bank of America, Innovative Scheduling, Yelp, Google, Oracle and Xerox - while the others are taking postdoctoral positions in the US and Switzerland. 

The doctoral research of these candidates has enhanced modelling, mathematics and computation in a wide variety of application areas including communications network design, scheduling, inventory management, sustainability, genomics, revenue management, public health, and financial markets.  

Some of the candidates will complete their degree work during the summer and will officially receive their degrees from Graduate School in August.   The graduation regalia of all was augmented at the ceremony with hoods placed by thesis advisors in recognition that students had now earned the status of scholarly colleagues.

Jiawei Qian has joined Bank of America / Merrill Lynch in Chicago as a quantitative strategist.  There, he designs trading strategies by solving large linear optimization problems in a few milliseconds.   His Ph.D. thesis, Prize-Collecting Network Design, deals with the design of fiber optic telecommunications networks.  The problem of designing and evolving these networks is known to be computationally hard to solve to an exact optimum for even the most powerful computers. 

Working with Professor David Williamson (right), Dr. Qian developed an approximation algorithm that handles the case in which there are both costs and revenues (hence "prize-collecting") associated with the network's links and termination points, some of which only become known over time as new service orders are received.  Born and raised in Beijing, Jiawei Qian earned his undergraduate degree in optimization and combinatorics and in computer science at the University of Waterloo in Canada.

Maurice Yuk Leung Cheung, who moved to Toronto from Hong Kong at the age of 12, also studied combinatorics and optimization at Waterloo.  He will work as a Senior Systems Engineer at Innovative Scheduling in Gainesville, Fla. 

Working with Professor David Shmoys (left), Dr. Cheung developed algorithms, based on linear programming, that find approximately optimal solutions to more general problems than previously had been solved well in the classical OR areas of machine scheduling and inventory management.  In both areas, his algorithms handle more realistic cost structures than are accommodated by existing well-solved models. 

Cheung's problems, like those of Qian and others, are known to be computationally hard.  Like Qian, he has proven that his solutions are within a specific factor of the true optimum.  Maurice Cheung's thesis is entitled LP-based Approximation Algorithms for Scheduling and Inventory Management Problems.  

Gwen Morgan Spencer grew up near Seattle, Wash. and graduated with a B.S. in mathematics from Harvey Mudd College in Claremont, Calif.  As a Thomas J. Watson Fellow, she travelled the world interviewing women scientists.  She is the first recipient of the Sherri Koenig Stuewer graduate fellowship, endowed by an ORIE alumna.  Spencer will be a Neukom Postdoctoral Fellow at Dartmouth College before joining the faculty of Smith College in 2014. 

With Professor Shmoys, Dr. Spencer formulated and developed approximately optimal solutions to models that capture key tradeoffs in containing the spread of wildfires (by using small-scale controlled burns) and of invasive species (by using biological control agents).  

Her models use networks to represent the spatial environments in which the wildfires and the invasive species spread in a random but predictable manner, and for which countermeasures have a predictable but uncertain impact.  Her work on these computationally hard problems can provide valuable decision support in the sustainable management of natural resources.  Gwen Spencer's thesis is entitled Approximation Algorithms for Stochastic Combinatorial Optimization, with Applications in Sustainability.

Scott Clark, from Tigard, Ore., graduated from Oregon State University with B.S. degrees in math, physics and computational physics.  At Cornell, he was enrolled in the Center for Applied Mathematics. 

Working with ORIE Professor Peter Frazier, Dr. Clark developed algorithms that work on a large number of computer processors concurrently and in parallel.  He analyzed the general problem of optimization using such parallel processing and applied this approach to problems in genomics, in a thesis entitled Parallel Machine Learning Algorithms in Bioinformatics and Global Optimization. 

One such problem is to simultaneously reassemble DNA code fragments from all of the genetic material collected from a community such as the bacteria in a person’s intestinal tract.  Clark likens this problem to reassembling a set of jigsaw puzzles from a collection of all the pieces from all of them – without the pictures on the boxes. His Ph.D. will be officially conferred in August. Scott Clark will work as a software engineer for Yelp, Inc. 

Chao Ding is from Luoyang, China, and has an undergraduate degree in Industrial Engineering from Tsinghua University in Beijing.  His research deals with the problem of deciding, as sales take place and data are acquired, whether or not to discontinue a product such as a cell phone that is sold under warranty and is subject to multiple failure types.  Each type of failure has a different repair cost and an unknown failure rate, which only becomes known with actual experience. 

With Professors Paat Rusmevichientong and Huseyin Topaloglu (who placed the hood), Ding constructed a model that uses automated learning to learn failure rates and detect unprofitable products in real time, and he came up with superior approximation methods to support the discontinuation decision.   His thesis is called High Dimensional Problems in Single Resource Revenue Management.  He expects his degree to be conferred in August. Chao Ding will work for Google in Mountain View, Calif. 

Kathleen Allison King is from St. Charles, Ill.  She is a graduate in computer engineering of the inaugural class at Franklin W. Olin College of Engineering in Needham, Mass.  She will work for Oracle in Boston.      

With Professor Jack Muckstadt, Dr. King developed optimization and simulation models to determine the best logistical design for distributing antibiotics to counter a large-scale anthrax attack. Her optimization and simulation models deal with such questions as determining the number of clinics needed to dispense medical supplies, how they should be staffed, and how and when supplies should be allocated to them.   Her results show the value of flexible clinic staffing plans and of centralized command and control in such an emergency. 

Using related models, King showed that the commercial pharmaceutical supply chain, rather than state public health departments, should be used to supply antiviral medication during an influenza pandemic.  This recommendation is under consideration by the Centers for Disease Control, which will use tools developed by Dr. King, Professor Muckstadt and ORIE M.Eng. graduate Christine Barnett to make decisions about the design and operation of the distribution system.   

Kathleen King's thesis is called Logistical Models for Planning and Operating Medical Countermeasure Distribution Networks during Public Health Emergencies.  Her degree will be officially conferred in August.

Martin Larsson is from Norrtälje, Sweden. After high school he studied Russian in the Swedish Armed Forces and worked as a translator at the Swedish Embassy in Moscow.  He then received an M.S. in engineering mathematics from Lund Institute of Techology in Sweden and subsequently a Master of Advanced Studies in finance from the Federal Institute of Technology in Zurich, Switzerland.  

In a thesis entitled Essays on the Mathematics of Market Efficiency, Dr. Larsson discusses the question of whether or not a financial market is efficient, i.e. whether market prices accurately reflect true value.   His is the first rigorous mathematical definition of efficiency, requiring that there be some economic equilibrium consistent with observed prices.  With this definition he developed a characterization of efficiency that can be tested by statistical means.   His advisors, who were unable to be present at the ceremony, are Professors Robert Jarrow and Sidney Resnick.  At right, he is congratulated by ORIE Director Adrian Lewis.

Martin Larsson's degree will be officially conferred in August.  He has accepted a postdoctoral fellowship at the École Polytechnique Fédérale in Lausanne, Switzerland. 

Juan Li is from Zhenjang, China, and received a B.A. in math at Smith College.  With Professor Muckstadt, Dr. Li worked on inventory management strategies applicable to the fulfillment of orders for a large on-line retailer such as Amazon.com.  Determining optimal inventory planning and execution strategies for each of tens of millions of products to be stocked at multiple warehouses and distribution centers that have limited capacity to store them is a daunting challenge.    She constructed mathematical models that describe the operations of the fulfillment system and developed and implemented computationally tractable algorithms for finding optimal stock levels.  

Juan Li's thesis is entitled Managing Inventory in a Large Scale Multi-echelon Capacitated Fulfillment System.  Her degree will be officially conferred in August.  She will join the Xerox Research Center in Webster, N.Y. At the May degree ceremony, the wedding of Juan Li and Hao Li on the previous day was recognized with the special banner held in the photo by Professors Muckstadt and Bland.