This REU site seeks to increase the number of students prepared to choose careers or graduate study in the computational sciences, bringing powerful computational, mathematical, and/or statistical techniques to bear on problems in science, engineering and education. It further seeks to give undergraduates with little opportunity to participate in research a chance to do so. Unique aspects of this REU site are:
- Interdisciplinary faculty mentors within one blended department; - A successful model based on internal pilot programs held in the summers of 2009 and 2010; -The on-site availability of cutting-edge research laboratories, including the NSF-supported MUGrid Project and Nexos Project, which were used during the pilot programs; - Inclusion of pre-service math teachers in research opportunities.
Throughout science and engineering fields, the availability of cheap computer processing power is driving increased interest in large-scale problems that were previously thought to be intractable through traditional methods. This can include developing new computational methodologies, software and/or systems, and new theoretical models. Predicting the movement of diseases, understanding the effects of large oil spills, forecasting consumer gas usage, and modeling the catastrophic effects of a volcanic eruption are just a few of the real problems that computers are being used to help solve. At the same time, ever smaller computing platforms have inserted computation into mobile, embedded, and ubiquitous realms of human experience that would have been unthinkable only a few years ago. Energy consumption of both massive computing clusters and battery-operated appliances is of great concern, and complex computer systems can now be found at the heart of everything from medical devices to the phones in most Americans' pockets. For both of these kinds of computational systems -- the big and the small -- there is a great need for practitioners with a deep understanding of computer science, modeling and statistical methods who can work across disciplines to build real, robust, and cost-effective solutions. This REU Site will bring 24 undergraduates to Marquette University over the next three years to be mentored by our interdisciplinary faculty of Mathematics, Statistics, and Computer Science in ways to approach the difficult computational problems that face our increasingly technological society.
Our mentor research expertise lies in all of these areas, as well as in how people learn in these fields, and how we prepare the next generation of students to explore computational sciences and mathematics. Each of the computational activities concerns students becoming involved in real applications using concepts described in the classroom. They will learn how to apply these abstract concepts to particular areas. Each application has been chosen so that it is accessible to undergraduates, yet is current and useful. Many research projects are smaller versions of potential Ph.D. dissertation topics, and each student works on their own portion of primary research. Our target audience is undergraduates of all levels majoring in mathematics, statistics, computer science, or related areas. The project includes a recruiting plan targeting a population of students with little or no access to research opportunities both within and outside of the immediate Milwaukee area. Underrepresented student populations will be targeted and encouraged to apply. Our mentor pool is highly diverse in both the academic and demographic senses, and we look forward to attracting some of the most talented students in the region and the nation for a summer of computational sciences.
This REU site seeks to increase the number of students prepared to choose careers or graduate study in the computational sciences, bringing powerful computational, mathematical, and/or statistical techniques to bear on problems in science, engineering and education. It further seeks to give undergraduates with little opportunity to participate in research a chance to do so. Across three summers (2011-2013), Marquette University's Department of Mathematics, Statistics and Computer Science hosted 26 NSF REU-sponsored students, plus an additional ten students from other funding sources, for a full-time, intensive research project with faculty mentors. Recruiting focused on colleges with little or no undergraduate research opportunities, or with significant underrepresented minority populations. Our popular site attracted more than ten times as many applicants than we were able to support. Participants demonstrated important attitudinal shifts about STEM research in subsequent long-term tracking. Participants went on to or plan to go on to graduate study in far greater numbers (50%) than the populations from which they were drawn. Two thirds of participants credited the REU program with giving them insight into or encouraging them to pursue graduate study. Student projects ranged from computational modeling to STEM education, and produced a variety of software releases and student co-authored research publications. One student's prototype improvements to GasDay Lab's popular forecasting tool resulted in a 30% improvement in forecast accuracy during testing. Better supply forecasts can ultimately translate to lower costs for natural gas customers. Each day, GasDay helps forecast about 20% of the natural gas delivered to residential, industrial, and commercial customers inthe US. Another student helped apply existing GasDay modeling techniques to a new resource venue -- residential water consumption. Her results identified ambient relative humidity as the key model parameter, and her model will help local startup company H2OScore.com to roll out a conservation tool designed to increase user awareness of current consumption patterns and offer solutions to motivate conservation in a time of widespread national drought and regional water source contention in southeastern Wisconsin. A third undergraduate researcher helped develop a new measure of "tortuousity" that may one day be helpful in gauging the health of blood vessel structures in the body. Students in the final summer of the program developed new software for allowing smartphones to intelligently determine when their user is "interruptable", and breathed new life into a popular software package for teaching hands-on, experimental computer science courses at institutions with limited laboratory resources. Other projects included explorations of grade school math teacher reasoning, smartphone collection of biometric data, power-aware computing, and computational modeling of ocean currents and aquatic ecosystems.
