Past Events

Seminars are held Tuesdays and Thursdays in POB 6.304 from 3:30-5:00 pm, unless otherwise noted. Speakers include scientists, researchers, visiting scholars, potential faculty, and ICES/UT Faculty or staff. Everyone is welcome to attend. Refreshments are served at 3:15 pm.

  • Additional Information

    Hosted by Leszek Demkowicz

    Sponsor: ICES Seminar

    Speaker: Mohamed Salah Ebeida

    Speaker Affiliation: Sandia National Laboratory

  • Abstract

    Bayesian inference provides a powerful framework for inductive reasoning which has proven applicable to an extensive range of disciplines and industrial applications. In this statistical theory, prior knowledge regarding quantities of interest are combined with observations to produce a full posterior distribution over the true values under consideration. MCMC methods have provided a practical framework for solving a very wide range of problems, yet there still remain a number of key difficulties and challenges for practitioners (e.g. issues with local trapping, inefficient mixing, correlated samples, and difficult convergence diagnostics).

    Our novel VoroSpokes framework for Bayesian inference resolves these issues by removing the underlying reliance on Markov Chains from the posterior sampling procedure. The posterior is instead adaptively approximated using an implicit domain partitioning in the form of Voronoi tessellations to construct a Voronoi Piecewise Surrogate (VPS) model. The surrogate model and domain partitioning are then used to prescribe a hierarchical sampling procedure designed to efficiently draw independent samples from the approximate posterior distribution. In this talk we will present the VoroSpokes framework, its theoretical properties and demonstrate its superior performance over MCMC in practice. ​​​

    Mohamed Ebeida is a research scientist at Sandia National Laboratories and an expert in computational geometry. He specializes in Voronoi diagrams, hyper-plane sampling and sphere packing. He developed a number of nontraditional Voronoi-based algorithms with application to meshing, uncertainty quantification, optimization, and machine learning. He graduated from University of California Davis in 2008 with a PhD in Mechanical and Aeronautical Engineering and a Masters in Applied Mathematics. He worked for two years as a Postdoc at Carnegie Mellon University. In 2010, he joined Sandia National Laboratories where he actively works in exploring the potential of implicit Voronoi tessellations for a wide range on applications in low and high-dimensions.

Tuesday, Jun 18

  • Additional Information

    Hosted by Fabrizio Bisetti

    Sponsor: ICES Seminar

    Speaker: Lucas Esclapez

    Speaker Affiliation: Lawrence Berkeley National Laboratory (LBNL)

  • Abstract

    Experiments have demonstrated that electric fields can be employed in combustion applications to control the flame stabilization and reduce pollutant emissions. However, the multi-scale nature of the physical processes has so far preventedthe development of efficient unsteady numerical methods, able to further analyze the interactions between turbulent flows,combustion kinetics and electric fields. We propose an algorithm for low Mach number combustion that incorporates the chemical production of charged species and their transport induced by electric field. The strategy relies on the multi-implicit spectral deferred correction (MISDC) approach that allows for a strong and efficient coupling of the stiff drift/diffusion and reaction terms with the slower advection terms. Along with the momentum, energy and species transport equations, a Poisson equation is solved to provide a consistent electrostatic potential. To overcome the stringent time-stepping constraint due to the fast electron motion, a non-linear implicit solve of the electrostatic potential and electron number density equations is used. For the strategy to be amendable to large scale computations, the non-linear system is solved using a Jacobian-Free Newton Krylov method, for which a preconditioner is developed. We will present the details of the method and demonstrate its capabilities on steady and unsteady of one-dimensional cases.

    This is joint work with J. B. Bell and M. S. Day.

    Dr. Lucas Esclapez is a post-doctoral researcher in the Center for Computational Sciences and Engineering, at Lawrence Berkeley National Laboratory (LBNL). His current research focuses on the development of high-performance simulation tools to study the interactions between flames and electric fields. Prior to his appointment at LBNL, he was a post-doctoral researcher at the Center for Turbulence Research (Stanford, USA. 2015-2016) and CERFACS (Toulouse, France. 2016-2018) working with Large Eddy Simulations (LES) to investigate the combustion process in industrial systems. Dr. Esclapez received is PhD from CERFACS in 2015 where he was studying aeronautical gas turbines ignition.

Tuesday, May 21

Identifying Actionable and Druggable Mutations from Cancer Big Data

Tuesday, May 21, 2019 from 3:30PM to 5PM | POB 6.304

  • Additional Information

    Hosted by Song (Stephen) Yi

    Sponsor: ICES Seminar

    Speaker: Zhongming Zhao

    Speaker Affiliation: Center for Precision Health, School of Biomedical Informatics and School of Public Health, UT Health Science Center, Houston, TX

  • Abstract

    As we have entered the precision medicine and big data science era, there are many unmet challenges on identifying the disease related information from large, heterogeneous data, and translating the findings for clinical use. Among these challenges, one is how to effectively identify driver mutations and genes in cancer genomes, especially those with the potential for druggable targets for the development of molecularly targeted cancer therapies. In this talk, I will introduce several informatics approaches, including SGDriver, AlloDriver, and KNMPx, to identifying cancer mutations and genes from large amount of somatic mutation data and our recently developed integrative network-based framework for identifying new druggable targets and anticancer indications from existing drugs.

    Dr. Zhongming Zhao holds Chair Professor for Precision Health and is the founding director of the Center for Precision Health, the University of Texas Health Science Center at Houston (UTHealth). Before he joined UTHealth in 2016, he was Ingram Endowed Professor of Cancer Research, Professor (with tenure) in the Departments of Biomedical Informatics, Psychiatry, and Cancer Biology of Vanderbilt University Medical Center, Chief Bioinformatics Officer of the Vanderbilt-Ingram Cancer Center (VICC), Director of the VICC Bioinformatics Resource Center, and the Associate Director of the Vanderbilt Center for Quantitative Sciences. Dr. Zhao has unique, interdisciplinary training: he received his master’s degrees in Genetics (1996), Biomathematics (1998), Computer Science (2002), Ph.D. degree in Human and Molecular Genetics (2000), and Postdoctoral Fellow in Bioinformatics (2001-2003). Dr. Zhao has broad interest in bioinformatics, genomics, precision medicine, and systems biology and has co-authored >330 papers in these areas. Dr. Zhao is the founding president of The International Association for Intelligent Biology and Medicine (IAIBM, 2018-).

Tuesday, May 14

Real-time error controlled surgical simulation: Towards personalised medicines

Tuesday, May 14, 2019 from 3:30PM to 5PM | POB 6.304

  • Additional Information

    Hosted by Tom Hughes

    Sponsor: ICES Seminar

    Speaker: Satyendra Tomar

    Speaker Affiliation: Head, Centre for Information Technologies & Applied Mathematics, U. Nova Gorica; Senior Researcher, Institute of Comp. Engin., U. Luxembourg

  • Abstract

    In this talk, the first a posteriori error-driven adaptive finite element approach for real-time simulation will be presented, and the method will be demonstrated on needle insertion problems. The used model is based on corotational elasticity and a frictional needle/tissue interaction. For simulating soft tissue deformation, the refinement strategy relies upon a hexahedron-based finite element method, combined with a posteriori error estimation driven local h-refinement. The local and global error levels in the mechanical fields (e.g., displacement or stresses) are controlled during the simulation. After showing the convergence of the algorithm on academic examples, its practical usability will be demonstrated on a percutaneous procedure involving needle insertion in a liver and brain. The brain shift phenomena is taken in to account which occurs when a craniotomy is performed.

    It is observed that the error in the computation of the displacement and stress fields is localised around the needle tip and the needle shaft during needle insertion simulation. By suitably and adaptively refining the mesh in this region, our approach enables to control, and thus to reduce, the error whilst maintaining a coarser mesh in other parts of the domain. Through academic and practical examples it will be demonstrated that our adaptive approach, as compared with a uniform coarse mesh, increases the accuracy of the displacement and stress fields around the needle shaft and, for a given accuracy, saves computational time with respect to a uniform finer mesh. This facilitates real-time simulations. Moreover, this work provides a first step to discriminate between discretization error and modeling error by providing a robust quantification of discretization error during simulations.

    The proposed methodology has direct implications in increasing the accuracy, and controlling the computational expense of the simulation of percutaneous procedures such as biopsy, brachytherapy, regional anaesthesia, or cryotherapy. Moreover, the proposed approach can be helpful in the development of robotic surgeries because the simulation taking place in the control loop of a robot needs to be accurate, and to occur in real time.

    The talk will conclude with some discussion on future outlook towards personalised medicines.

    After completing his Ph.D. from IIT Kanpur (on parallel h-p spectral element methods), he worked as a postdoctoral researcher on numerical simulation of water waves using discontinuous Galerkin method at University of Twente, Netherlands. Thereafter, from 2005 to 2014, he worked as a research scientist at RICAM, Austria, and developed optimal order iterative solvers, and reliable and sharp a posteriori estimates for non-conforming approximations. Since 2015, as a senior researcher at University of Luxembourg, he is working on real-time surgical simulations and some meshless/meshfree methods. Moreover, since October 2018, he is also leading the Centre of Information Technologies and Applied Mathematics at University of Nova Gorica, Slovenia.

Friday, May 10

Student Blitz Presentations

Friday, May 10, 2019 from 11AM to 12PM | POB 6.304

  • Additional Information

    Hosted by Max Bremer and William Ruys

    Sponsor: ICES Seminar-Student Forum Series

    Speaker: Meghana Palukuri, Joshua Chen, Tim Smith

    Speaker Affiliation: CSEM Program, Oden Institute, UT Austin

  • Abstract

    Student Blitz presentations are a series of short technical presentations. The aim of these talks is to improve awareness of other students' research and foster intra-departmental collaboration. This weeks talks are:
    (1) "A supervised machine learning approach to community detection in networks." Meghana Palukuri
    (2) "Determining local cluster metric tensors for local dimension reduction." Josh Chen
    (3) "SAMBA Dances to a Tropical Pacific Rhythm." Tim Smith

Friday, May 3

Randomized algorithms for accelerating matrix computations

Friday, May 3, 2019 from 10AM to 11AM | POB 6.304

  • Additional Information

    Hosted by Tom O'Leary-Roseberry and Kendrick Shepherd

    Sponsor: ICES Seminar-Babuska Forum Series

    Speaker: Per-Gunnar J. Martinsson

    Speaker Affiliation: Professor, Mathematics & Moncrief Endowed Chair No. 4 in Simulation-Based Engineering Sciences, Oden Institute, UT Austin

  • Abstract

    Low-rank matrix approximations, such as partial spectral decompositions or principal component analysis (PCA), play a central role in data analysis and scientific computing. The talk will describe a set of randomized algorithms for efficiently computing such approximations. These techniques exploit modern computational architectures more fully than classical methods and enable many computations involving massive data sets.

    The algorithms described are supported by a rigorous mathematical analysis that exploits recent work in random matrix theory. The talk will briefly review some of the key theoretical results.

Thursday, May 2

Modeling, dynamics, and control of wall-bounded shear flows

Thursday, May 2, 2019 from 3:30PM to 5PM | POB 2.402 (Electronic)

  • Additional Information

    Hosted by Takashi Tanaka

    Sponsor: ICES Seminar

    Speaker: Mihailo Jovanovic

    Speaker Affiliation: Professor, University of Southern California

  • Abstract

    Understanding and controlling transitional and turbulent flows is one of the most important problems in fluid mechanics. In the first part of the talk, techniques from control theory are used to examine the early stages of transition in wall-bounded shear flows. We demonstrate high sensitivity of the flow equations to modeling imperfections and show that control theory can be used not only to design flow control algorithms but also to provide valuable insights into the transition mechanisms.

    In the second part of the talk, we describe how to account for second-order statistics of turbulent flows using low-complexity stochastic dynamical models based on the linearized Navier-Stokes (NS) equations. The complexity is quantified by the number of degrees of freedom in the linearized evolution model that are directly influenced by stochastic excitation sources. For the case where only a subset of correlations are known, we develop a framework to complete unavailable second-order statistics in a way that is consistent with linearization around turbulent mean velocity. In general, white-in-time stochastic forcing is not sufficient to explain turbulent flow statistics. We develop models for colored-in-time forcing using a maximum entropy formulation together with a regularization that serves as a proxy for rank minimization. We show that colored-in-time excitation of the NS equations can also be interpreted as a low-rank modification to the generator of the linearized dynamics. Our method provides a data-driven refinement of models that originate from first principles and it captures complex dynamics of turbulent flows in a way that is tractable for analysis, optimization, and control design.

    Mihailo R. Jovanovic is a professor in the Ming Hsieh Department of Electrical and Computer Engineering and the founding director of the Center for Systems and Control at the University of Southern California. He was a faculty in the Department of Electrical and Computer Engineering at the University of Minnesota, Minneapolis, from December 2004 until January 2017, and has held visiting positions with Stanford University and the Institute for Mathematics and its Applications. His current research focuses on dynamics and control of fluid flows, large-scale and distributed optimization, design of controller architectures, and fundamental limitations in the control of large networks of dynamical systems. He serves as an Associate Editor of the IEEE Transactions on Control of Network Systems, and had served as the Chair of the APS External Affairs Committee, a Program Vice-Chair of the 55th IEEE Conference on Decision and Control, an Associate Editor of the SIAM Journal on Control and Optimization (from 2014 until 2017), and an Associate Editor of the IEEE Control Systems Society Conference Editorial Board (from 2006 until 2010). Prof. Jovanovic is a fellow of APS and IEEE. He received a CAREER Award from the National Science Foundation in 2007, the George S. Axelby Outstanding Paper Award from the IEEE Control Systems Society in 2013, and the Distinguished Alumnus Award from the Mechanical Engineering Department at UC Santa Barbara in 2014.

Tuesday, Apr 30

Ensemble Kalman Inversion: from subsurface environments to composite materials

Tuesday, Apr 30, 2019 from 3:30PM to 5PM | POB 6.304

  • Additional Information

    Hosted by Clint Dawson

    Sponsor: ICES Seminar

    Speaker: Marco Iglesias Hernandez

    Speaker Affiliation: University of Nottingham, United Kingdom

  • Abstract

    The Ensemble Kalman filter (EnKF) developed by Evensen and co-workers in the 1990s has had enormous impact in the geosciences and various engineering disciplines. EnKF has been historically used for data assimilation problems, where the objective is to infer the state of a partially observed dynamic system from observational data. Motivated by algorithmic ideas in EnKF, Ensemble Kalman Inversion (EKI) is a computational framework that has been recently proposed for solving PDE-constrained inverse problems (i.e. to infer inputs from outputs of a PDE model) in a derivative-free fashion. In this talk I will introduce EKI from a framework that unifies both the Bayesian and the classical (deterministic) approach for inverse problems. I will present regularisation strategies for EKI that can improve accuracy and performance of large-scale inversions. I will further discuss recent parameterisations within EKI which enable to efficiently infer geometric features of the underlying (unknown) field. Numerical examples will be used to show the potential advantages of these parameterisations in various application areas including the non-destructive evaluation of composite materials as well as the geoelectrical characterisation of the subsurface.

Tuesday, Apr 30

Unlock the Personal Sky - Safe and Assured Autonomy for On-Demand Urban Air Mobility

Tuesday, Apr 30, 2019 from 11AM to 12PM | POB 6.304

Important Update: NOTE: Different time
  • Additional Information

    Hosted by Ufuk Topcu

    Sponsor: ICES Seminar

    Speaker: Peng Wei

    Speaker Affiliation: Assistant Professor, Department of Aerospace Engineering, Iowa State University

  • Abstract

    Urban Air Mobility (UAM) is an envisioned air transportation concept, where intelligent flying machines could safely and efficiently transport passengers and cargo within urban areas by rising above traffic congestion on the ground. Companies such as Boeing, Airbus, Bell, Embraer, Joby, Zee Aero, Pipistrel, and Volocopter are working with their battery vendors to build and test electric vertical takeoff and landing (eVTOL) aircraft to ensure that vehicle safety and energy efficiency become an integral part of people’s daily commute. Furthermore, in order to make UAM profitable for operators and affordable for passengers, the flight operations must be able to scale, which means that the expected air traffic density will be extremely high. For example, as one of the industry leaders in UAM, Uber estimated more than 5,000 eVTOL flights per day in the city of Los Angeles for its future scaled Uber Air operations. The UAM community recognized a key challenge remaining unanswered to make UAM a reality: how can we design and build a real-time, trustworthy, safety-critical autonomous UAM ecosystem to enable large-scale flight operations in high-density, dynamic and complex urban airspace environments? In this talk the speaker will present preliminary studies to address this critical research challenge from areas in autonomy, control, real-time systems and safety. Our multidisciplinary approach is based on bridging guidance and control, reinforcement learning, and Markov decision process.

    Peng Wei is an assistant professor in Iowa State University Aerospace Engineering Department, with courtesy appointments in Electrical and Computer Engineering Department and Computer Science Department. Prof. Wei is leading the Intelligent Aerospace Systems Lab (IASL). By contributing to the intersection of control, optimization, machine learning, and artificial intelligence, he designs autonomous and human-in-the-loop decision making systems for aeronautics, aviation and aerial robotics. Recent applications include: Air Traffic Control/Management (ATC/M), Airline Operations, UAS Traffic Management (UTM), eVTOL Urban Air Mobility (UAM) and Autonomous Drone Racing (ADR). Prof. Wei received his undergraduate degree in Information Science and Control Theory from Tsinghua University and a Ph.D. degree in Aerospace Engineering from Purdue University. He serves in several advisory boards at Airbus and NASA. He is an associate editor of AIAA Journal of Aerospace Information Systems.

Monday, Apr 29

  • Additional Information

    Hosted by Dmitrii Makarov

    Sponsor: ICES Seminar-Molecular Biophysics Series

    Speaker: Igor Rubstov

    Speaker Affiliation: Tulane University

  • Abstract

    We developed small size vibrational labels capable of testing membrane mobility via two-dimensional infrared spectroscopy. Despite low polarity of the membrane interior, the vibrational label shows a significant inhomogeneous linewidth, which enables measuring accurately lipid membrane mobilities via time-resolved spectral diffusion technique. Because of its small size, the label can be anchored to a specific depth in a bilayer, assessing its local mobility.