Mathematics of Data Science, Machine Learning and Optimization
Researchers in the mathematics of data science, machine learning, and optimization focus on algorithmic complexity, combinatorial optimization, and machine learning applications, including deep learning for partial differential equations and discrete optimization.
Related People
Biography
George Turkiyyah is a research professor in the Applied Mathematics and Computational Science program at KAUST.
Before joining KAUST, he was a professor at the American University of Beirut, where he also served as chair of the computer science department. Prior to joining AUB, he was an assistant professor and later an associate professor at the University of Washington in Seattle.
Turkiyyah earned a Bachelor of Engineering (B.Eng.) in civil and environmental engineering from the American University of Beirut, and both a Master of Science (M.S.) and a Doctor of Philosophy (Ph.D.) in computer-aided engineering from Carnegie Mellon University.
Turkiyyah has been involved in the development of knowledge-based AI systems that have been deployed in practice. He has also developed several widely used simulation codes for high-resolution finite element engineering applications. His work on fast methods for surgical simulation has led to a software startup and several patents.
His research has earned several awards, including the Transportation Research Board K.B. Woods Award in 2003 for best paper in design and construction, the Best Presentation Award at the ACM Solid and Physical Modeling Conference in 2007, and the Best Poster Award at the Medicine Meets Virtual Reality Conference in 2006.
He chaired the 2003 ASCE Engineering Mechanics Conference and co-chaired the Eighth ACM Symposium on Solid Modeling and Applications (SPM) in 2003. He is a member of ACM and the Society for Industrial and Applied Mathematics (SIAM).
Research Interests
Professor Turkiyyah’s current research interests include hierarchically low-rank matrix algorithms and their HPC/GPU implementations to support the development of simulation models at extreme scales.
His work addresses various applications of hierarchical matrix technology, including PDE-constrained optimization, high-dimensional statistics problems, multi-dimensional fractional diffusion problems, scientific data compression and second-order methods for training neural networks.
Education
Biography
Mikhail Moshkov is a professor of Applied Mathematics and Computational Science (AMCS) and an affiliated professor of Computer Science (CS) at KAUST. He is also the principal investigator of the Extensions of Dynamic Programming, Machine Learning, Discrete Optimization (TREES) research group.
Professor Moshkov holds an M.S. summa cum laude in 1977 from the University of Nizhni Novgorod, Russia. He obtained his Ph.D. in 1983 from the University of Saratov, Russia, and a Doctor of Science in 1999 from Moscow State University, Russia.
Before joining KAUST, he held professorships at the University of Nizhni Novgorod, Russia, and the University of Silesia, Poland.
Moshkov received the State Scientific Stipend in Mathematics for Outstanding Scientists from April 2000 to March 2003, awarded by the Presidium of the Russian Academy of Sciences. Additionally, he received the First Degree Research Prize, awarded by the rector of the University of Silesia, Poland, in 2006.
Research Interests
Professor Moshkov's research interests include: (i) The study of time complexity of algorithms in computational models such as decision trees, decision rule systems and acyclic programs with applications to combinatorial optimization, fault diagnosis, pattern recognition, machine learning, data mining, and analysis of Bayesian networks. (ii) The analysis and design of classifiers based on decision trees, reducts, decision rule systems, inhibitory rule systems, and lazy learning algorithms. (iii) Extensions of dynamic programming for sequential optimization relative to different cost functions and for study of relationships between two cost functions with applications to combinatorial optimization and data mining.
Rolf Krause
applied mathematics Numerical Solution of Partial Differential Equations Finite elements machine learning Numerical Optimization medicine computational mechanics contact problems fluid-structure interactions cardiac simulation biomechanics geology Multiphysics Simulation HPC optimization Multigrid Domain Decomposition software development
Biography
Rolf Krause is a full professor in the Applied Mathematics and Computational Sciences Program at KAUST, with a career spanning academia, research and leadership. Before joining KAUST, he was a full professor at Università della Svizzera italiana in Lugano, Switzerland, where he directed the Institute of Computational Science from 2009 to 2020 and has served as co-director of the Center for Computational Medicine in Cardiology since 2014.
Beyond his research roles, Professor Krause has held notable leadership positions including a director of the interdisciplinary Euler Institute in 2021 and was the founding dean of the Faculty of Mathematics and Informatics at UniDistance Switzerland in 2022.
His commitment to academic service includes roles such as chairman of the examination board for mathematics studies and chairman of the board of finance for tuition fees at the University of Bonn from 2007 to 2009, as well as membership in the academic senate at USI from 2017 to 2021.
His work has earned numerous awards, including the Taylor & Francis Prize for Innovative Contribution to Theoretical Biomechanics/Biomedical Engineering and the MATH+ Distinguished Visiting Scholar recognition from the MATH+ Center in Berlin.
Professor Krause holds a Doctor rerum naturalium in Mathematics with distinction ("summa cum laude") from The Free University of Berlin, awarded in 2001. He also earned a Diploma in Mathematics with a minor in Economics from the same institution in 1996.
Research Interests
Professor Krause's research focuses on numerical simulation, machine learning, optimization, and data-driven approaches. A major focus of his research is the design and analysis of efficient and reliable algorithms that can be used to solve complex problems in scientific computing and machine learning.
Krause and his colleagues use mathematical understanding and computer science expertise to advance sustainable progress in many areas, from medicine to geology. They provide scientific software capable of solving complex, large-scale problems that can run on modern supercomputers such as KAUST’s Shaheen III.
Areas of expertise and focus
- Contact problems in mechanics
- Scientific software
- Multilevel and domain decomposition methods
- Optimization
- Iterative solution of large-scale systems
- Parallel computing
- High-performance computing (HPC)
- Coupled problems
- Finite elements
- Non-linear solution methods
- Neural networks
- Physics-informed neural networks
- Cardiac simulation
- Biomechanics
- Computational geoscience
Application areas
- Medicine
- Computational mechanics
- Contact problems
- Fluid-structure interaction
- Cardiac simulation
- Biomechanics
- Geology
- Complex and coupled multiphysics
Education
Biography
Professor Tempone received his Ph.D. in Numerical Analysis ('02) from the Royal Institute of Technology, Sweden. The next phase of his career took him to the United States, where he completed his postdoctoral studies at the University of Texas' Institute for Computational and Engineering Sciences (ICES), before joining Florida State University as an Assistant Professor of Mathematics.
Tempone joined KAUST in 2009 as a founding faculty, with the rank of Associate Professor of Applied Mathematics before becoming a Full Professor of Applied Mathematics in 2015. He is also the principal investigator of the Stochastics Numerics Research Group at KAUST.
A variety of fields, such as computational mechanics, quantitative finance, biological and chemical modelling and wireless communications, are driving his research. More specifically, his research contributions include a posteriori error approximation and related adaptive algorithms for numerical solutions to deterministic and stochastic differential equations. His honors include the German Alexander von Humboldt professorship (2018-2025) and the first Dahlquist Fellowship in Sweden (2007-2008). He was elected Program Director of the SIAM Uncertainty Quantification Activity Group (2013-2014).
Research Interests
Professor Raul Tempone's expertise and research interests lie at the intersection of applied mathematics, computational science, and stochastic analysis, with a strong focus on developing and analyzing numerical methods for stochastic and deterministic problems. His work emphasizes adaptive algorithms, Bayesian inverse problems, scientific machine learning, stochastic optimization, and uncertainty quantification, aiming to push the boundaries of computational efficiency and accuracy in simulations.
At the helm of the Stochastic Numerics Research Group at KAUST, Tempone is particularly interested in applications spanning computational mechanics, quantitative finance, biological and chemical modeling, and wireless communications. His research group is dedicated to tackling a posteriori error approximation, data assimilation, hierarchical and sparse approximation, optimal control, optimal experimental design, and the rigorous analysis of numerical methods.
Professor Tempone's approach is not only theoretical but also highly applicable, seeking to address real-world problems in various domains by leveraging mathematical and computational techniques. His work is instrumental for those interested in the practical application of mathematics to solve complex, real-world issues, making his research group an ideal place for potential collaborators, postgraduate students, postdocs, and research scientists looking for cutting-edge projects at the nexus of uncertainty quantification and computational science.