SAMSI hosted the Trends and Advances in Monte Carlo Sampling Algorithms Workshop, part of the Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applied Mathematics (QMC) Program on the campus of Duke University from Dec. 11-15, 2017.
The workshop was attended by more than 100 experts in the fields of applied mathematics, statistics and machine learning for the purpose of exchanging ideas and advancing the broad area of sampling algorithms.
This event was the second workshop presented in the QMC program and featured how Monte Carlo sampling methods can be used to help optimize performance of machines and/or business and industrial processes. This complex methodology is widely used in physics, chemistry, mathematics and statistics, and is most useful when other methods fail due to the high dimensionality of the problem.
Participants enjoyed a week-long workshop that featured talks from innovative mathematicians from around the world. The talks focused on research being done in the field of Monte Carlo sampling and how these applications can be used to tackle real-world problems in business and industry.
The QMC program has ten working groups that were created in the QMC Opening Workshop in late August 2017. The working groups support research being done by applied mathematicians, statisticians and researchers across a wide variety of topics. The working groups will re-convene at the QMC Transition Workshop in May 2018 to discuss their findings and to develop collaborations between colleagues for future research.
The Trends and Advances Workshop is one of many ways in which SAMSI continues to promote the importance of applied mathematics, statistics and computational science. To see the research presented, visit the workshop webpage at: https://www.samsi.info/qmc-trends-and-advances.
In January 2018, SAMSI welcomed its third director, David Banks, a Professor of the Practice of Statistics from Duke University’s Department of Statistical Science.
“SAMSI is amazing…I’ve been involved since 2003, and I have watched it grow and evolve,” said Banks upon being announced as the new director.
Banks took over the position from Richard Smith, the Mark L. Reed III Distinguished Professor of Statistics and Professor of Biostatistics from the University of North Carolina at Chapel Hill’s Department of Statistics and Operations Research (STOR). Smith has served as SAMSI Director since 2010 and now assumes the role of an Associate Director at SAMSI.
During his tenure, Smith did a great deal to enhance the SAMSI brand by working to bring in interesting programs that highlighted the importance of statistics and applied mathematics across a broad spectrum of subjects. From forensic science to astronomy or computational methods for large data and climate research, Smith worked with the SAMSI directorate and staff to bring in fresh programs organized by some of the leading experts in their fields from around the world. In his new role as an associate director, Smith will focus more of his efforts towards his passion of teaching and climate research.
Banks obtained his Master of Science in Applied Mathematics from the Virginia Polytechnic Institute and State University in 1982, followed by a Ph.D at the same school in Statistics in 1984. In his career, Banks has served in numerous academic institutions and government organizations. One of Banks’ most prestigious positions was as Chief Statistician of the U.S. Department of Transportation in the late 1990’s, followed by a stint at the U.S. Food and Drug Administration in 2002. Banks returned to academics in 2003, where he joined the Department of Statistical Science at Duke.
“Every time you change jobs you get a new skill set, a new set of friends, some new ideas and a raise,” said Banks. “If you change jobs well, you keep the old friends, skills and thinking. Changing jobs is positive, and I hope my move to SAMSI will be as gratifying has my previous job hopping has been.”
In addition to his many professional accomplishments, Banks has also written scholarly papers and has served as an editor of the Journal of the American Statistical Association, as well as co-founding the journal of Statistics and Public Policy, where he also served as an editor. He has also published 74 refereed articles, edited eight books, and co-authored four monographs.
In his research, Banks enjoys statistical modeling the most because the research offers insight into the explanations of complex problems. His research areas also include models for dynamic networks, dynamic text networks, adversarial risk analysis (i.e., Bayesian behavioral game theory), human rights statistics, agent-based models, forensics, and certain topics in high-dimensional data analysis.
Banks recently served as the president of the International Society for Business and Industrial Statistics. He is a fellow of the American Statistical Association and of the Institute of Mathematical Statistics. He also won the American Statistical Association’s Founders Award in 2015.
Banks’ said that for now, until he gets more comfortable in his new position as director at SAMSI, he will focus research goals towards data science and machine learning methodologies. Everyone at SAMSI welcomes Banks as the new director and looks forward to working with him.
After many months SAMSI is proud to welcome their newest Deputy Director, Elvan Ceyhan.
Ceyhan, who was a visiting associate professor at the University of Pittsburgh in 2016, joined the SAMSI directorate in July this year. He joins the SAMSI team and will also serve as a research associate professor of the Department of Statistics at North Carolina State University (NCSU). He replaces former Deputy Director, Sujit Ghosh, who is currently a Professor of Statistics in the same department at NCSU.
Ceyhan, a Turkish native, received his undergraduate education and a Bachelor of Science in Mathematics from Koc University (KU) in Istanbul, Turkey. In 1997, he came to the United States and originally attended Oklahoma State University’s (OSU) Ph.D. Program in Mathematics, before changing his mind and switching to their statistics master’s program. He went on to receive his Master of Science degree in Statistics from OSU in 2000. That same year, Ceyhan began the Ph.D. program in the Applied Mathematics and Statistics Department at Johns Hopkins University – he went on to receive his Ph.D. from Johns Hopkins in 2004.
From 2004 to 2005, Ceyhan worked as a postdoctoral fellow at the Center for Imaging Science at Johns Hopkins. After his time at Johns Hopkins, he returned to Turkey and served as an assistant professor in the Department of Mathematics at KU until 2011, when he was promoted to an associate professorship. Ceyhan served in that capacity until 2016, when he went to the University of Pittsburgh for the visiting associate professor post. Throughout his academic career, he has (co) authored almost 50 journal articles and given numerous talks and presentations.
What Ceyhan enjoys most about applied mathematics and statistics is data analysis, finding hidden patterns and studying trends in data. “I was always good in math in primary school,” he said. “I entered the university as a physics major and a year later, realized I liked math better, so I switched.”
Ceyhan is easy going and enjoys working with members of the SAMSI directorate, the staff and postdoctoral fellows and visitors that attend the institute. After Elvan took the position, he decided the best thing to do was to pick up where his predecessors had left off in order to increase awareness of how SAMSI supports applied math and statistics fields.
“I would like to continue our conventions and contribute more effort towards diversity in our programs,” said Ceyhan. He also believes SAMSI needs to continue to support heavily data science and big data programs, as these topics are major points of interest in the statistics community.
“I would like to continue our conventions and contribute more effort towards diversity in our programs,” said Ceyhan. He also believes SAMSI needs to continue to support heavily data science and big data programs, as these topics are major points of interest in the statistics community.
Among his many goals as deputy director, Ceyhan will work to expand education and outreach initiatives, support undergraduate workshops and programs and serve as an advisor to postdoctoral fellows in order to help them advance their research and academic careers.
Ceyhan resides with his wife, of nearly 10 years and his two children, daughter Gokce and son Melih. His family moved with him in 2016 when he took the visiting associate professor position in Pittsburgh. The family still misses Turkey and they hope to get back to the country next year to visit.
Ceyhan enjoys watching soccer and studying ancient history in his spare time. SAMSI is glad to have him in this new leadership role within the organization.
Since she began her academic career, Leslie McClure has always had a keen interest and respect for others. It is her passion for representing women and people of color in the mathematical sciences that led to her recently being appointed as the SAMSI Associate Director of Diversity in early August of this year.
“Throughout my education, particularly my undergraduate, I was often one of very few women in my classes, and rarely had female professors,” said McClure. “Women and people of color are represented in lower numbers in the professorate, and have even less representation in the higher levels of academics.”
Before working at Drexel, McClure spent 11 years as a faculty member at the University of Alabama at Birmingham in the Department of Biostatistics.
“I gravitated towards biostatistics because it was a good fit with my interests, but I think I was also attracted to the field because it appeared more diverse than math,” said McClure.
McClure is a trained clinical trials statistician and her current research is focused on the methods that drive adaptive design in clinical trials, as well as the practical implications of implementing an adaptive design. She also works diligently trying to understand why racial inequalities exist in disease, particularly cardiovascular disease and stroke, and the role that the environment may play in those differences.
“Without diversity of people, we do not have diversity of ideas. Without diversity of ideas, we lose creativity in science, and fail to continue moving forward,” — Leslie McClure
Much like her research, McClure also works as a champion to find ways to make the field of mathematics more inclusive to women and under-represented minorities. “As I have pursued my own academic goals, I have also worked to increase and maintain diversity in the math sciences,” she said.
McClure is also part of the leadership for the National Alliance for Doctoral Studies in the Mathematical Sciences, where she serves as the Associate Director of Statistics. One of the main goals of the Math Alliance, located on the campus of Purdue University, is to foster the growth of the community of mathematical scientists in order to promote a diverse workforce.
SAMSI is proud to have added such an accomplished professional to the directorate. SAMSI believes whole-heartedly in creating an academic environment of equality and inclusivity for all. As the SAMSI Diversity Director, McClure will work with local universities and through her numerous contacts nationwide to research and implement strategies that will work towards advancing the careers of under-represented groups in the field mathematics.
McClure stays busy and focused. When she is not doing research or working for the betterment of others, she stays active running and spending time with her husband (who is a Chemistry Professor at Drexel) and their two children, Lillian and Preston. McClure also enjoys spending time with her dog, Bosco, watching Law and Order reruns.
“Without diversity of people, we do not have diversity of ideas. Without diversity of ideas, we lose creativity in science, and fail to continue moving forward,” said McClure.
Late August was indeed a busy time for SAMSI as the opening workshops occurred in succession during the end of that month. The week-long CLIM Opening Workshop ran from Aug. 21-25 and the QMC Opening Workshop (Aug. 28 – Sept. 1) served as the starting point for both programs.
The CLIM Program looks at analyzing data and climate models to potentially predict future changes on our Earth that could directly impact our environment and the human population. The CLIM Opening Workshop featured many esteemed minds in the study environmental science. The opening workshop led to the creation of 13 working groups, whose overall purpose is to study various factors and data analysis in order to understand how our environment is evolving.
“Climate Science is important for many reasons in our society,” said Richard Smith, Director of SAMSI and Leader of the CLIM Program about the opening workshop. “It is not widely appreciated just how critical the role of mathematical and statistical methods play in climate science.”
More than 120 participants from universities around the world attended the popular workshop. Twenty-five speakers presented lectures on various topics about the science of the environment and how to use mathematical and statistical data to find the root to the causality seen in the modern world. The two panel discussions held during the workshop created much discussion and offered many contributions that led to the creation of the CLIM Program working groups.
The workshop participants were even treated to a rare solar eclipse that occurred over the continental United States during that time. To accommodate this rare event, organizers planned time during the opening day to go out and view the phenomena as it reached the totality phase. Everyone was excited as they used solar eclipse glasses and/or various safe methods to view the eclipse. The last time a solar eclipse could be viewed from the contiguous United States was Feb. 29, 1979. The eclipse was a special occurrence that was a happy coincidence to fall during the workshop and offered a perspective of how much we are shaped by the world around us.
As the opening workshop closed, participants chose the working groups they would be affiliated with for the remainder of the CLIM program. The workshop created valuable network opportunities between the scientists and mathematicians in attendance so that they can continue their research even after the CLIM program ends in May next year.
“This workshop brought together some of the top experts in climate science with the leading researchers in mathematics and statistics,” said Smith. “The lively discussions generated many ideas that will be developed during the rest of this [CLIM] program.”
The QMC Opening Workshop began the following week, Monday, Aug. 28, and was hosted at the beautiful Penn Pavilion on the campus of Duke University.
This workshop brought together more than 110 mathematicians and statisticians, who collectively created 10 specific working groups focused on discussing ways in which they would research how to use big data across a wide range of practical applications.
“Kudos to the QMC Program Leaders Art Owen, Frances Kuo, Fred Hickernell and Pierre L’Ecuyer for getting the year-long SAMSI QMC off to a fantastic start,” said Ilse Ipsen, Associate Director of SAMSI and the QMC Program Leader. “Their commitment, combined with spot-on real-time assistance from SAMSI postdocs Cheng Cheng, Matthias Sachs and Whitney Huang, produced this lively Opening Workshop and an unusually large number of 10 promising working groups.”
The goal of the QMC Program is to explore the potential of QMC and other deterministic, randomized and hybrid sampling methods for a wide range of applications, including the numerical solution of PDEs; machine learning; computer graphics; Markov chain sampling, like MCMC and MCQMC; sequential Monte Carlo; and uncertainty quantification.
More than 20 speakers were invited to speak on a wide variety of sampling methods. The talks generated much discussion amongst participants and created the impetus for the working groups that were created.
Overall, the QMC Opening Workshop was well received by the participants and many looked forward to the future meetings in their respective working groups.
“The QMC Program is well on its way to being super-productive,” said Ipsen.
The research that will come from both the CLIM and QMC programs will help to address ways in which we can improve our environment, improve efficiency and productivity through random sampling across various applications, and advance technology. Research and collaboration are how SAMSI works to advance research in statistics and applied mathematics to innovate the future.
Cheng is in the SAMSI QMC Program. She earned her Ph.D. in Mathematics from the University of Central Florida in 2017. Her research interests include applied and computational harmonic analysis, emphasis on sampling theory in signal processing, and high dimensional data analysis.
Yawen is in the SAMSI CLIM Program. She received her Ph.D. in statistics from Pennsylvania State University. During her graduate studies, she was fortunate to work with top scientists in the study of the Antarctic Ice Sheet. She was intrigued by the ice sheet physics and developed a statistical method to combine physics and multiple data sets to study ice streams on West Antarctica. Her research interests are spatial statistics, Bayesian modeling and computational methods for large data.
Huang is in the SAMSI CLIM Program. He received his Ph.D. in Statistics from King Abdullah University of Science & Technology (KAUST). His research experiences include computational methods for spatio-temporal statistics and functional data analysis. He enjoys using these statistical tools to collaborate with other scientists who have expertise in climate, oceanography, geophysics, etc., in order to explore interesting environmental problems.
Whitney is in the SAMSI CLIM Program. His research focuses on statistics of extremes and spatial, spatio-temporal data analysis with applications in climate. Ultimately, his research goal, as a statistician, is to bridge the gap between statistics and atmospheric/oceanic sciences. In his spare time he enjoys hiking, travel, and watching sports (basketball, tennis).
Maggie is in the SAMSI CLIM Program. She received her Ph.D. in Statistics from Iowa State University in 2017. Her broad research interests are in developing statistical methods for solving environmental problems. Some of her particular statistical research interests are in temporal and spatiotemporal statistics, Bayesian statistics, hierarchical modeling, and mixture models. She is originally from Minnesota where her family owns a Highland cattle farm, and in her spare time she enjoys cooking, fly-fishing, and woodcarving.
Mikael is in the SAMSI CLIM Program. He is a statistician working on data analysis methods for physical science applications. He is currently working on developing spatio-temporal interpolation techniques for analysis of oceanographic data from Argo profiling floats. In his free time, he enjoys traveling to far-away places, hiking in the summer and skiing in the winter.
Matthias is in the SAMSI QMC Program. During his Ph.D. he has have been working on numerical methods for ergodic stochastic differential equations. He has focused his efforts towards working on discretization methods for variants of the Langevin equation with applications in canonical sampling and molecular modelling. He is currently exploring the application of these models in sampling problems in Bayesian statistics and machine learning.
Christian is in the SAMSI CLIM Program. He just received his Ph.D. in mathematics from the University of Utah this year. His research interests lie at the interface between geophysics and mathematics. He is interested in sea ice and its role in the Earth’s climate system. While at SAMSI, he will be working with Professor Chris Jones at UNC Chapel Hill.
The IMSM is an annual educational outreach event that features collaborations with industry, national labs and other governmental organizations. During the workshop graduate students in mathematics, statistics and computational science disciplines are exposed to challenging real-world problems that arise in industrial and government laboratory research.
“This type of summer workshop has been held at N.C. State since 1995,” said Mansoor Haider, a Professor of Mathematics at N.C. State University and the workshop’s organizer. “The IMSM name has been in place for well over a decade now. This reflects the importance of integrating statistics with mathematics and computation in solving modeling problems arising outside of academia.”
Several prominent leaders in industry and national labs provided first-hand experience and mentorship to the students. This year SAMSI was proud to partner with professionals from: Sandia National Laboratories; Rho, Inc.; U.S. Army Corps of Engineers, PAREXEL and the Environmental Protection Agency (EPA) among others.
This year’s partners presented problems to the attending students. The students were placed into research groups, and then collectively developed and implemented ways to resolve the issues at hand. The various partner representatives and workshop faculty members provided valuable mentorship and direction to the students. The students also received practical experience in problem-solving and first-hand experience in what it is like to work in a research group in a non-academic setting.
“The need for doctorally trained statisticians and mathematicians in industry and national labs is ever increasing,” said Haider. “By immersing them in an intensive collaborative research experience, we hope to increase students’ awareness of the variety of career options after graduation, and the skills they will need to be successful.”
Some of the problems tackled by this year’s participants included:
How to integrate large-scale data from open source Google Earth Engine with air quality monitoring across the country in order to provide real-time air quality information to users.
Using coast line bathymetry data to assist in erosion control – to be used in predicting environmental effects after coastal storms or helping humanitarian aid logisticians to identify effective delivery methods by sea to provide critical relief.
Determining root causes of allergies in humans by studying the correlation and interactions between microbes in the environment and those inside the nose. Potential applications of this research aim to adapt the design of buildings and control exposure to identified allergens to reduce allergy and asthma among children at risk.
Participants put in long hours, sometimes well into the night. The students worked together and maximized the individual knowledge strengths of group participants to assist in solving their team’s assigned problem. After several days of team research and collaboration with group mentors and faculty, the groups reconvened and presented their findings to their peers and other academic professionals. The partners attending the workshop got valuable responses to the problems they posed, and in some instances, received insight into alternative research avenues or approaches to pursue in the future.
“For many students, this is their first experience tackling mathematical or statistical modeling problems outside of a university research setting,” said Haider. “The workshop is intensive… It nicely mimics unique challenges in industrial research like identifying, formulating, and solving problems in a team, and then refining, coordinating, presenting and reporting on the results, all in a short time period.”
The IMSM is one of the many ways SAMSI helps bring new talent together in order to collaborate with relevant applied math, statistics and computational science organizations. These workshops help to prepare and inspire those considering careers in science and math disciplines for the future.
For more on the Industrial Math/Stat Modeling Workshop and to see research presented from previous workshops, visit: www.samsi.info/imsm-history.
Nearly 40 participants attended the workshop in order to discuss their findings compiled from multiple working groups formed throughout the past academic year. The organizing committee for the program listened to spokespersons from each group as they presented their findings. The group also discussed continuing future collaborations between these working groups once the program was over.
SAMSI’s ASTRO program liaison and Deputy Director, Sujit Ghosh, noted that the ASTRO program has been successful in creating a cohesive bond between the statistical and mathematical sciences and the disciplinary sciences, like astronomy and astrophysics. According to Ghosh, this coupling is helping to systematically streamline the analysis of huge data sets that are produced from the Laser Interferometer Gravitational-Wave Observatory (LIGO), gravitational wave (GW) research and exoplanet discoveries.
A panel, consisting of ASTRO program leaders, collected feedback from the numerous researchers in attendance. A significant issue that researchers brought up was the challenge of publishing research articles in domain sciences (i.e. core stat or astrophysics journals) versus the disciplinary sciences. This issue was viewed as a significant obstacle when using these research papers as a reference for tenure-based decisions. The panel of program leaders could not determine the best way to address this situation. Instead they agreed that this topic should be readdressed during future interdisciplinary engagements, like transition workshops.
“This [ICTS-SAMSI] workshop helped form several collaborations to enable what will likely prove to be a fruitful collaboration among people from diverse backgrounds that can propel the progress of science”
Overall, the ASTRO Program focused on ways to create solid partnerships between researchers in applied mathematics, astronomy, astrophysics and statistics (professionals who do not ordinarily work together in the field). In fact, the concept of astrostatistics emerged from numerous collaborations, like this one, between researchers during past SAMSI programs. The partnerships created by this program are important because they could potentially advance research in astronomy. In addition, three mid-program workshops (one on Exoplanets in the Fall of 2016 and two on Synoptic Surveys and GW Astronomy and Astrophysical Population Emulation in the Spring of 2017) were organized by the researchers to support the program during the past year.
Bengaluru, India. This workshop enabled scientists to share their ideas and work together across two continents in order to explore the grand challenges in gravitational waves time domain astronomy.
“SAMSI workshops and working groups have helped me understand how my thesis work fits into the larger scientific picture and how to gain a better understanding of what our science priorities are as a community of observational astronomers,” said Jackeline Moreno, a graduate student at Drexel University, who was a member of one of the working groups. Moreno said she was impressed with how the joint workshop brought together experts from around the world and from different research backgrounds to come together and share techniques and insights for analyzing time series data.
Kaustabh Vaghmare, a data scientist from the Inter -University Center for Astronomy and Astrophysics (IUCAA) in Pune, India, who also attended the ICTS-SAMSI workshop, agreed. Vaghmare began by saying that time domain astronomy has improved a great deal in the last decade, due in large part, to advances in robotic telescopes, image processing and database technologies. These advances, according to Vaghmare, have given astronomers the ability to organize several systematic surveys of the sky. In addition to those advances though, Vaghmare sited the importance of the human aspect as a valuable way of sharing information. “This [ICTS-SAMSI] workshop helped form several collaborations to enable what will likely prove to be a fruitful collaboration among people from diverse backgrounds that can propel the progress of science,” he said.
Joint workshops, like the ICTS-SAMSI workshop, help SAMSI to emphasize the value of collaborating with other institutions or across fields of study. The results of these collaborations creates more dynamic ways to solve traditional problems using the tools of applied mathematics and statistics as a guide.
The Optimization Transition Workshop was hosted by SAMSI from May 1-3, 2017 to effectively close the Program on Optimization for the 2016-2017 research year.
The workshop was attended by nearly 40 participants who discussed findings in research conducted from the program’s 13 different working groups formed last fall. Scholars and researchers from multiple fields of applied math and statistical science not only explored progress made by their groups throughout this past year, but they also discussed effective ways to collaborate after the program was over in hopes of continuing to tackle some of the complex issues the field of optimization presents.
Overall, the program was attended by some 500 participants throughout this past research year. The program offered an opportunity for colleagues to share their knowledge and to network with up and coming researchers in the field. Programs like this one help SAMSI to provide researchers in statistics, applied mathematics and data science fields a forum to meet, discuss and collaborate on a wide range of topics.
SAMSI is one of eight math institutes funded by the National Science Foundation (NSF) whose purpose is to advance research in the mathematical sciences, increase awareness of mathematical sciences and disciplines and directly engage prospective intellectual talent in that effort.
SAMSI hosted an Undergraduate Workshop as part of its Education and Outreach initiative in their Program on Optimization from Feb. 27-28, 2017.
Nearly 40 undergraduate students from universities across the country were treated to lectures on optimization methods used in large-scale statistical analysis and were also introduced to statistical inverse problems. In addition, students received hands on familiarization with software packages that help to determine these complex calculations.
Though the workshop only lasted two days, students stayed busy! They received an overview on who SAMSI is and how they are helping to support and promote those considering the fields of mathematics and statistics. Students also met and networked with SAMSI Postdoctoral Fellows. Post docs mentored the young group on what they should focus on in their academics to get ready for the job market in the fields of applied mathematics and statistics.
On the last day, students took a field trip to SAS, a major internationally known software company, headquartered in N.C.
During their visit, the students received talks from computer scientists, analytical mathematicians and optimization specialists on how SAS develops software in line with client user and business objectives.