Math-to-Industry Boot Camp X

Overview

The Math-to-Industry Boot Camp is an intense six-week session designed to provide graduate students with training and experience that is valuable for employment outside of academia. The program is targeted at Ph.D. students in pure and applied mathematics. The boot camp consists of courses in the basics of programming, data analysis, and mathematical modeling. Students work in teams on projects and are provided with training in resume and interview preparation as well as teamwork.

There are two group projects during the session: a small-scale project designed to introduce the concept of solving open-ended problems and working in teams, and a "capstone project" that is posed by industrial scientists. Recent industrial sponsors included Cargill, Securian Financial and CH Robinson. Weekly seminars by speakers from many industry sectors provide the students with opportunities to learn about a variety of possible future careers.

Eligibility & Logistics

Applicants must be current graduate students in a Ph.D. program at a U.S. institution during the period of the boot camp. The program will take place online. Students will receive a $3,000 stipend.

Organizers

Thomas Hoft, University of St. Thomas
Daniel Spirn, University of Minnesota, Twin Cities

2025 Boot Camp Projects

Mortgage-backed security price modeling – aka, the 750 billion dollar question

US Bank, Christopher Jones, Marshall Smith

A typical commercial bank deploys 20-25% of its assets into some form of a security, which is typically a mortgage-backed security (MBS) or U.S. Treasury security. In aggregate, the securities held by banks totals $5.5 Trillion as of May 2025. As part of accounting regulations, unrealized gains/losses impact a bank’s income statement. And depending on the bank’s size, it also impacts its capital. Therefore, understanding price movements of securities is of high importance. Historically, banks have preferred MBS because these securities produce higher yields than Treasury securities. However, this situation has changed following the interest rate moves experienced since 2022. At any given time there is a range of coupons (the annual income of the security) that trade at various prices. MBS prices have fluctuated substantially over the past 5 years due to Federal Reserve intervention in the financial markets during the COVID-19 pandemic and subsequently due to the higher interest rate environment that began in 2022. In this project, students will explore a model for the representation of prices of MBS relative to the coupon of the securities in question. The conclusion of this project should include a model for explaining price movement of MBS.

Ebenezer Acquah, University of Tennessee
Eduardo Torres Davila, University of Minnesota
Chanuka Dissanayake, University of Pittsburgh
Alex Parker, Iowa State University
Buzheng Shan, Texas A&M University

Proactive Housing Support: Using Public Data to Predict and Prevent Evictions

Hennepin County, Liz Sprangel

Human services provides critical support to residents facing housing instability, and lacks reliable predictive tools to identify those most at risk of eviction. Using Hennepin County eviction rates and socioeconomic factors, this project aims to build predictive models of eviction filings across neighborhoods. We will utilize advanced analytics and statistics to estimate eviction risk at the census tract level, and forecast the overall eviction rates in the county. This project will highlight unique challenges of working with sensitive data, including addressing concerns of equity and evaluating social impact.

Deborpita Biswas, Clemson University
Baboucarr Dibba, University of Texas Rio Grande
Maruf Lawal, University of Tennessee
Jan Schmidt, University of Nebraska

Modeling Cell-Cell Interactions in T-Cell Dependent Cytotoxicity (TDCC) Based Cancer Therapeutics

Takeda Pharmaceutical, Dean Bottino

Cancer research has recently focused on exploiting T cell (more generally, effector cell) capabilities to detect and kill cancer cells. Cell therapies directly deliver engineered T cells against tumor-associated antigens (TAAs) to patients, while T-cell engagers facilitate synapse formation between T-cells and TAA-positive cells. In both cases, initial discovery and optimization of these therapies rely heavily on in vitro assays which measure the degree of tumor (and healthy) cell killing as a function of the effector-to-target (E:T) cell ratio initially seeded into each well. It is not at all clear (to the mathematical modeling world, at least) that killing is truly invariant under E:T if we proportionally increase both the numerator and the denominator. The goal of the boot camp project team (BCPT) will be to use mathematical modeling to determine the proper functional relationship between effector and target cell densities and cytotoxicity, under in vitro conditions (typically a layer of target cells at the bottom of the well, with effector cells added to the medium above the target cells). We envision this project will initially employ agent-based modeling to develop our intuition, potentially followed by stochastic PDEs and more advanced modeling techniques, ultimately resulting in adequate ODE approximations to in vitro effector-target interactions. Time and resource permitting, we will also model real experimental data to iteratively test and refine the models.

Parnian (Maedeh) Ahmadzadeh, University of Houston
Idowu Esther Ijaodoro, University of Alabama
Jenita Jahangir, University of Louisiana at Lafayette
Zedan Liu, University of Miami
Emeka Mazi, Georgia State University

Sensor data and Chinese economic activity

Cargill, Kaisa Taipale, Anup Singh

Utilize air pollution data as an alternate indicator for economic activity in emerging markets, especially China. Economic activity creates air pollution, via power generation, biomass burning, transport, and manufacturing, and so may be a viable feature for modeling economic activity. Base load power generation in many emerging markets is still heavily fueled by coal, and marginal power generation is often from natural gas. The project is to use publicly available sensor data on air pollution and other data as desired in combination with economic indicators from China to create a robust model for economic activity across different timescales. Complexities will include economic indicators that don’t always agree; somewhat messy publicly available datasets; geospatial processing; and trends stemming from the energy transition and evolution of manufacturing in China. Intermediate steps will require acquiring and organizing the publicly available data from disparate sources to serve project goals. This is a good project for those interested in time series, geospatial data, economics, geopolitics, or the energy transition. While Python would likely be the preferred language, R has some excellent libraries for geospatial analysis and would be appropriate as well.

Daniel Arreola, Iowa State University
Rick Danner, University of Vermont
Joe Macula, University of Colorado
Chi Nguyen, Virginia Tech
Sehong Park, University of Miami