Math-to-Industry Boot Camp V

• Mentor Marshall Lagani, Securian Financial
• Gabrielle Angeloro, Iowa State University
• Adeyemi Fagbade, University of Wyoming
• Priyanga Ganesan, Texas A & M University
• Chang Li, University of Central Florida
• Liban Mohamed, University of Wisconsin, Madison
• Radmir Sultamuratov, Wayne State University
• Jidong Wang, University of Oregon

It has become commonplace for insurance companies to offer products that link benefit guarantees to stock market indices, such as the S&P 500. Modeling the risks inherent in such a product requires a strong understanding of mathematical finance as well as significant computational resources. Derivatives instruments, primarily futures, options, and swaps, can be used to hedge the liability, providing an effective mitigation of product risks.
Participants will learn about variable annuities, a common equity-linked product, as well as some of the common derivatives instruments used to hedge the risks in these products. We will explore some of the techniques used to model the liabilities they generate and develop methods to create proxy models, allowing us to monitor risks and rebalance hedge positions intraday as the markets move in between model runs. This project assumes little to no background in mathematical finance and should be of interest to participants who are interested in computational statistics, quantitative finance, and Python.

• Mentor Nicole Bridgland, World Wide Technology
• Cameron Cook, University of Tennessee
• Erica de la Canal, The University of Texas at Austin
• Kari Eifler, Texas A & M University
• Nazar Emirov, University of Central Florida
• Hansen Pei, University of Delaware (Newark, DE, US)
• John Portin, University of Kansas
• Katherine Weber, University of Minnesota, Twin Cities

Supply chain operations motivate many data science and optimization problems. From a demand and pricing perspective, one might ask: how much of item X do we anticipate selling? How much do we expect it to pay for it, depending on when we buy it? From a storage and operations perspective, one might ask how we best store it in warehouses to get it to where it's going. Do we have enough warehouse space for all the stuff we will need to store in the near future? What are the error bars on that space usage estimate? There's plenty of questions from a purely operational perspective as well. For example, in a busy warehouse, forklift traffic can cause significant slowdowns. A forklift at one load or drop-off location may block access to several locations in the warehouse. Forklifts waiting to enter one row could block the major paths through the warehouse. This project is directed at optimizing internal warehouse transit operations, through any of storage location choices, job scheduling, or pathing choices.

• Mentor Yu-Li Huang, Mayo Clinic
• Nawaf Alansari, The Pennsylvania State University
• Skye Binegar, Georgia Institute of Technology
• Ryan Coopergard, University of Minnesota, Twin Cities
• Alketa Henderson, University of North Carolina, Greensboro
• Dhir Patel, The Ohio State University
• Yanru Su, University of Kansas
• Zhimin Wu, Arizona State University

Bone Marrow Transplant (BMT) is an effective treatment for many hematological malignancies. This modality has become integral to the management of many patients resulting in a dramatic increase in the volume of patients undergoing the procedure. The volume of patients coming for transplant (about 500 patients undergo this highly complex procedure annually at Mayo Clinic Rochester) has progressively increased over the past decade leading to many innovative solutions to adapt to this challenge. Over the past two decades the infrastructure has been developed to allow a majority of patients to undergo many components of the procedure as an outpatient visit despite the highly complex nature of the patients and associated risk of complications. Ultimately we have reached maximum safe capacity with our current workflow. This has posed major stresses on many areas including patient scheduling, stem cell collection, outpatient visit, human cellular therapy laboratory, hospital based outpatient facility, and inpatient facility. BMT practice has recently implemented a predictive model for stem cell collections. This model is expected to increase capacity by 20% with the same resources. The practice also adopted pre scheduling concept to plan for entire patient transplant itinerary starting from stem cell collections, pre-chemo visits, to chemo treatment and stem cell infusion. There are uncertainties in all three stages due to patient conditions, resource constraints, and process complexity. This short term project will focus on modeling and optimizing the stochastic nature of these three stages which could potentially provide recommendations for scheduling policy and resource planning.