Applied Mathematical and Computational Sciences

Program in
Applied Mathematical
and Computational Sciences

Graduate
College

Monday, November 23 2009 @ 01:26 AM CST

AMCS Seminar

Spring 2008

February 1 - Professor Gregory R. Carmichael, Department of Chemical and Biochemical Engineering, "Predicting Air Quality: Improvements through advanced methods to integrate models and measurements".
Abstract: Air quality prediction plays an important role in the management of our environment. Computational power and efficiencies have advanced to the point where chemical transport models can predict pollution in an urban air shed with spatial resolution less than a kilometer, and cover the globe with a horizontal resolution of less than 50 km. Predicting air quality remains a challenge due to the complexity of the governing processes and the strong coupling across scales. While air quality prediction is closely aligned with weather prediction, there are important differences, including the role of pollution emissions and their associated large uncertainties. Improvements in air quality prediction require a close integration of observations. As more atmospheric chemical observations become available chemical data assimilation is expected to play an essential role in air quality forecasting. In this seminar advances in air quality forecasting are discussed with an emphasis on data assimilation. Applications of the four dimensional variational method (4D-Var) and the ensemble Kalman Filter (EnKF) approach are presented and the computation challenges are discussed. Read More
February 8 - Cancelled due to the following joint Math and AMCS Colloquium talk by Dr. Yongcheng Zhou, Center for Theoretical Biological Physics, University of California, San Diego, "Continuum modeling of macromolecular electrostatics and conformational change: mathematical analysis and simulations".
Abstract: The conformational charge in large spatial and temporal scale is essential for the proper functions of macromolecules such as ion channel gating, membrane transpass and molecular motors; Such large deformation usually involve the electrostatic force as the major driving force. This talk will be focused on the continuum modeling and computation of the electrostatic force and its coupling with the large macromolecular deformation modeled as nonlinear elastic displacement. We will investigate the regularity and the solvability of these electro-elasticity equations, analyze a stable regularization method for solving the singular electrostatic potential from the Poisson-Boltzmann equation, discuss the numerical issues related to harmonic extension of the elastic displacement in the solvent region, and apply this model to an important biological process: the BAR domain-induced membrane curvature. I will also highlight the mathematical problems for refining this model and other ongoing projects about the electro-diffusion in molecular and cellular levels. Read More
February 15 - Professor Ying Zhang, Department of Biostatistics, "Semiparametric Analysis of Panel Count Data".
Abstract: Abstract Read More
February 22 - Professor Samuel A. Burer, Department of Management Sciences, "An Introduction to Semidefinite Programming and Its Applications".
Abstract: Over the past 15 years, the field of semidefinite programming -- which fits somewhere between linear programming and general convex programming -- has grown tremendously. This growth can be attributed to a wealth of applications in areas such as combinatorial optimization, machine learning, and probability theory. In this tutorial, we introduce semidefinite programming, trace its history, touch on algorithms and software for its solution, and detail a number of applications. Read More
February 29 - Professor Jasbir S. Arora, Department of Civil and Environmental Engineering, "Optimization-based Motion Prediction of Mechanical Systems: Sensitivity Analysis".
Abstract: In this study, we derive sensitivity equations for the problem of optimization-based motion prediction of a mechanical system using the inverse recursive Lagrangian formulation. The simulation and sensitivity formulations are based on Denavit-Hartenberg transformation matrices. External forces and moments are taken into account in the formulation. The sensitivity information is needed in the optimization-based simulation process. The proposed formulation is demonstrated by calculating sensitivities for the optimal time trajectory planning problem of a two-link manipulator. In addition, sensitivities obtained using the proposed algorithm are compared to those obtained using the closed-form equations of motion. The two sensitivities match quite closely. The lifting motion of the two-link manipulator with external loads is also optimized by using the algorithm developed in this paper. Applications of the proposed formulation to digital human motion prediction are presented and demonstrated. Read More
March 7 - Dr. Jerome Pansera, Geode Capital Management, Boston, "Consistent interest-rate modeling".
Abstract: Interest rates are dependent on two variables: the calendar time and the time to maturity. They can thus be represented as a surface. In this talk, we discuss two pitfalls that may occur when building a mathematical model of this surface: arbitrage opportunities and model inconsistency. We also present one of the simplest models that avoid these two pitfalls. Finally, we estimate the parameters of this simple model.
March 14 - Professor William Klink, Department of Physics and Astronomy, "Symmetry and Relativistic Nuclear Physics".
Abstract: Nuclear forces can be thought of as arising in two different ways. On the one hand nucleons (protons and neutrons) interact via the exchange of mesons, such as pi mesons. On this view the nuclear forces that bind nucleons to produce nuclei are the result of meson exchange. On the other hand both nucleons and mesons are thought to be made out of more fundamental entities called quarks. Quarks interact via particles called gluons to produce the known nucleons and nuclei. Underlying both of these viewpoints is a theoretical framework called quantum field theory. A great deal of work is going on to find suitable approximations and numerical techniques for solving the equations of quantum field theories. In this talk I will describe work being done with Kevin Murphy, who is in the AMCS program, that deals with diagonalizing large matrices, the low lying eigenvalues of which give the observed nuclear spectrum. Read More
March 28 - Mr. Kamran Kazmi, AMCS, "Bioluminescence tomography with optimized optical parameters".
Abstract: Bioluminescence tomography (BLT) is a rapidly developing new area of molecular imaging that allows in vivo studies of ongoing biological processes at cellular and molecular levels. In bioluminescence imaging, animal organs and tissues are labeled with bioluminescent reporter probes that generate detectable signals. The goal of BLT is to produce a quantitative reconstruction of the bioluminescent source distribution within the living animal from bioluminescent signals measured at the surface of the body. While in most BLT studies so far the optical parameters of the key anatomical regions are assumed to be known from the literature or diffuse optical tomography, these parameters cannot be very accurate in general. In this talk, we propose a new BLT approach that optimizes optical parameters when an underlying bioluminescent source distribution is reconstructed to match the measured data. Read More
April 4 - Professor Teodor Rus, Department of Computer Science, "Liberate computer user from programming".
Abstract: Computer programming is error-prone, tedious, and difficult. It is error-prone because the programmer need to handle machine logic; it is tedious because it is repetitive; and it is difficult because it requires the programmer to be both, a problem-domain expert (to develop the algorithms) and a computer expert (to represent algorithms in machine-language). To alleviate these impediments software tools have been developed. However, over the short life of computers we witnessed a tremendous increase in the amount and diversity of problems solved by computers. The result is a huge increase in software tool sophistication and complexity; the side-effect is a proportional increase in computer-based problem solving difficulty. What is to be done to prevent this? All solutions proposed to fight software complexity are based on the development of new-programming languages. Since new programming languages mean new software-tools this only lead to more software complexity! What we actually need are software tools that would make computers understand natural language of the problem domain as the domain expert does. Such software tools would actually liberate computer user from programming! In this talk I will discuss a new computer-based problems solving methodology that provides steps toward computer user liberation from programming. Read More
April 11 - Professor Rodica Curtu, Department of Mathematics, "Frequency control in neuronal competition models".
Abstract: The notion of competition has been applied in the modeling of various neuronal computational tasks. For example, competition between two neuronal populations, via reciprocal inhibition, and slow adaptation underlie models for alternating rhythmic behavior in central pattern generators (CPGs) and in perceptual bistability. CPGs consist of neural circuits that drive alternately contracting muscle groups. Perceptual bistability refers to a class of phenomena in which a deeply ambiguous stimulus gives rise to two different interpretations that alternate over time, only one being perceived at any given moment. In both cases the question is what mechanism underlies the oscillation?s frequency (or period) that is believed to encoding information about the stimulus to the neural system. In this talk we focus on a class of competition firing rate models and show that the period of oscillations depends non-monotonically on the stimulus strength. The period increases with input strength for low values of the latter and in this case the dynamics is due to a release mechanism. On the other hand, for large values of input strength the period decreases with it and the dynamics is due to an escape mechanism. Read More
April 18 - Professor Sharif Rahman, Department of Mechanical and Industrial Engineering, "Random Eigenvalue Problems in Computational Dynamics".
Abstract: I will present a new class of computational methods, referred to as dimensional decomposition methods, for solving a general random eigenvalue problem in modeling and simulation of stochastic dynamic systems. The methods involve a novel function decomposition that facilitates lower-dimensional approximations of a general multivariate function, Lagrange interpolation of lower-variate functions, and Monte Carlo simulation. Compared with the commonly used perturbation methods, neither the assumption of small input variability nor the calculation of partial derivatives of eigensolutions is required by methods developed. Due to a small number of original function evaluations, the proposed methods are effective, particularly when an eigenvalue analysis entails costly finite-element, mesh-free, or other numerical analysis. Numerical examples from linear dynamics will be presented to evaluate the accuracy, convergence, and computational efficiency of the methods. If time permits, an industrial-scale, complex-valued eigenvalue problem involving brake squeal analysis will also be discussed. Read More
April 25 - Professor Michael Lovaglia, Department of Sociology, "Two Unsolved Problems in Mathematical Sociology: Status in Task Groups and Power in Exchange Networks".
Abstract: Two sociological research problems have been refined to the point where they are particularly amenable to mathematical analysis. The first problem involves the relative status of individuals working together on a group task. Influence is a key indicator of status rank in a task group. The problem is to predict the relative influence of various group members based on status characteristics known to alter influence: gender, race, age, education, beauty, height, and occupation. Of particular interest is the relative impact of gender and race. For example, in a four-person work group composed of an African-American woman, an African-American man, a European-American woman, and a European-American man, predict the amount of influence each will have over group decisions. The second problem involves the relative power of positions in an exchange network to acquire resources in competition with others. Exchange is defined as a transaction profitable to both parties to it. An exchange network is similar to a market except that not all sellers have access to all buyers. The links that allow exchange between particular positions form networks of infinite variety. In simplest form A-B signifies that A and B can exchange. A-B-C signifies that A and B can exchange, B and C can exchange, but A and C cannot exchange. A common exchange rule is that positions can exchange at most once during a given round of exchange. Thus B might exchange with A or C but not both. Over a series of exchange opportunities (rounds), B will acquire nearly all the resources available as A and C compete for the single available exchange opportunity with B. Adding positions and links changes the relative power of positions to successfully acquire resources. Given the shape of the network and the rules of exchange, predict the amount of resources that each position in the network will acquire. Current attempts to solve both problems use graphical analysis. Read More
May 2 - Professor Elias S.W. Shiu, Department of Statistics and Actuarial Science, "Dynamic Fund Protection".
Abstract: Nowadays, many products sold by life insurance companies are basically mutual funds, wrapped around with some insurance features or guarantees. These guarantees are financial options that should be priced, hedged, and reserved using modern option-pricing theory, with mathematical tools such as martingales, Brownian motion, stochastic differential equations, and so on. This talk will discuss the product “dynamic fund protection” and it is partially based on the paper “Pricing perpetual fund protection with withdrawal option” http://www.soa.org/library/journals/north-american-actuarial-journal/2003/april/naaj0304-4.pdf Read More
May 9 - Professor Anthony Paik, Department of Sociology, "Social Networks: A Sociological Perspective". Read More

Spring 2007

January 26 - Palle Jorgensen, Mathematics Department, University of Iowa, "Some uses of mathematics of wavelets in signal/image processing".
Abstract: We will survey some parallel developments in engineering and in mathematics which have proved of mutual benefit to both subjects. The mathematics led to better algorithms, used for example by the FBI in finger printing, and in medical imaging. Signal processing in turn was used in an essential way in the mathematical developments of wavelets. Read More
February 2 - Weimin Han, Mathematics Department, University of Iowa, "Numerical Analysis in Iowa". Read More
Friday, February 9 - Jia Lu, Department of Mechanical and Industrial Engineering, University of Iowa, "Inverse Elastostatic Problem in Biomechanics".
Abstract: In this talk, we will discuss some recent developments on some inverse problems in biomechanics. Most biological structures are permanently loaded in they service life. Their in vivo shapes correspond to deformed states and hence, it is the deformed states that are typically given as the starting point of biomechanical analysis. This brings both challenge and opportunities for mechanics. For example, in stress analysis, it is imperative to use methods that will correctly account for the pre-existing deformations. First, we will present an inverse formulation of finite element method designated to solve elastostatic problem of pre-deformed bodies. The method exists on an inverted paradigm of mechanics, where the equilibrium problem is solved by means of finding the initial geometry. Then, we will discuss some unique advantages of the inverse approach in the context of membrane problems. We will show that the membrane stress can be practically determined independent of material model. Thirdly, we will introduce some on-going works along the line of utilizing inverse method to characterize the elastic properties of biological membranes.
February 16 - Bruce Ayati, Department of Mathematics, Southern Methodist University, "A Multiscale Model of Biofilm as a Senescence-Structured Fluid".
Abstract: We derive a physiologically structured multiscale model for biofilm development. The model has components on two spatial scales, which induce different time scales into the problem. The macroscopic behavior of the system is modeled using growth-induced flow in a domain with a moving boundary. Cell-level processes are incorporated into the model using a so-called physiologically structured variable to represent cell senescence, which in turn affects cell division and mortality. We present computational results for our models which shed light on modeling the combined role senescence and the biofilm state play in the defense strategy of bacteria.
February 23 - Wayne Polyzou, Department of Physics and Astronomy, University of Iowa, "Relativistic Three Nucleon Scattering".
Abstract: Our eyes see objects by processing the light rays scattered by the object. The size of the smallest object that can be seen is limited by the wavelength of the light. Elementary particles are about 8 orders of magnitude smaller than the wavelength of light. Physicists ``see'' elementary particles by measuring the reaction products of collisions at distances far from the point of collision. The minimum size that can be resolved is limited by the de Broglie wavelength of the incident particle, which can be made as small as desired by increasing the energy of the incident particle. In order to resolve structures on sub-nuclear scales (less than $10^$ meters) the velocities of the colliding particles must approach the speed of light. The theory that is used to study these study these collisions is quantum scattering theory. In this talk I will discuss (1) the formulation of quantum scattering theory (1) what can be learned from quantum scattering theory (2) the mathematical formulation of the equations for three three particle scattering (3) recent work on relativistic three particle scattering.
March 2 - Osnat Stramer, Department of Statistics and Actuarial Science, University of Iowa, "Inference for Partially Observed Diffusion Processes: A Critical Review".
Abstract: Diffusion processes are a well established tool for modeling continuous time data from a wide range of areas. The development of statistical inference for diffusion processes is thus of high practical importance. We will evaluate and compare some of the different methods to estimate the parameters of a diffusion process. We will emphasis justification and practical implementation of various techniques. Finally we will point out some interesting future projects.
March 9 - Alberto Maria Segre, Department of Computer Science, University of Iowa, "Grappling with Protein Structure Prediction".
Abstract: One of the grand challenges of computational biology is to determine a protein's three dimensional shape (or ``tertiary structure'') -- and, consequently, provide insight into its biological function -- from its molecular composition (or ``primary structure''), expressed as the sequence of constituent amino acids. This ``protein structure prediction problem'' can be seen as an enormously difficult optimization problem, both because of the number of possible configurations a protein might assume and because we don't yet precisely understand the underlying energetics. In this talk, I will describe our hybrid optimization approach to the structure prediction problem. Our approach blends a distributed branch-and-bound algorithm with a number of secondary search techniques, including some based on continuous optimization techniques.
March 30 - Ian Besse, Department of Mathematics, University of Iowa, "Math Models in Electrophysiology".
Abstract: This talk will focus on the development of mathematical models which describe the electrophysiology of so called excitable cells, a class that includes neurons, cardiac myocytes and skeletal muscle fibers. I will begin with an introduction to the physiological processes underlying the "excitability" of such cells, in particular those processes related to the generation of electrical impulses in neurons. This will be followed by a discussion of the characteristics that any accurate model of these processes must exhibit and how one particularly famous model, the Hodgkin-Huxley ODE model, was been able to capture these characteristics so well that even today it still remains the gold standard for models within computational neurophysiology.
April 6 - Hantao Zhang, Department of Computer Sciences, University of Iowa, "Finding Small Combinatorial Designs by a SAT solver".
Abstract: Combinatorial design theory has extensive interactions with numerous areas of mathematics and applications in numerous disciplines. It is also a rich source of structured, parameterized families of hard propositional theories for SAT solvers, which decide if a propositional theory has models. As the language of the first NP-complete problem, propositional logic is expressive enough to encode small size problems in combinatorial design theory. Recent dramatic improvements in the efficiency of SAT solvers encourage people to attack combinatorial design problems by encoding problems as propositional formulas, and then searching for their models using off-the-shelf general purpose SAT solvers. In this talk, we will present a case study of such attempts involving the existences of latin squares with certain properties. We show how to encode a latin square problem efficiently in propositional logic and how to use SAT solvers effectively. This study helps us to not only gain new knowledge about design theory but also provide insight to improve the performance of SAT solvers.
April 20 - Roberto A. Saenz, Department of Applied Maths and Theoretical Physics Centre for Mathematical Sciences, Cambridge University, "A within-host model of influenza virus".
Abstract: Influenza virus is a major threat to public health, causing millions of cases around the world every year. Although there is a good understanding of the epidemiology of the influenza virus, viral kinetics have been given limited attention. Quantitative information on viral kinetics is crucial in assessing the appropriateness of intervention measures. Using a simple mathematical model, we intend to give a description of the dynamics of the influenza virus inside an infected host and its interaction with the host's immune system.
April 27 - Thomas L. Casavant, Depts. of Electrical and Computer, Biomedical Engineering, and Ophthalmology, University of Iowa, "Grappling with Protein Structure Prediction".
Abstract: This talk describes and evaluates a coarse-grained parallel computational approach to identifying rare evolutionary events often referred to as "horizontal gene transfers". Unlike classical genetic evolution, in which variations in genes accumulate gradually within and among species, horizontal transfer events result in a set of potentially important genes which "jump" directly from the genetic material of one species to another. Such genes, known as xenologs, appear as anomalies when phylogenetic trees are compared for normal and xenologous genes from the same sets of species. However, this has not been previously possible due to a lack of data and computational capacity. With the availability of large numbers of network-connected compute clusters, as well as genomic sequence from more than 2,000 species containing as many as 35,000 genes each, and trillions of sequence nucleotides in all, the possibility exists to examine clusters of genes using phylogenetic tree similarity as a distance metric. The full version of this problem requires years of CPU time, yet only makes modest IPC and memory demands; thus, it is an ideal candidate for a grid computing approach. This paper describes the prototype of such a solution and preliminary benchmarking results that show a reduction in total execution time from approximately two years to less than one day. I will also report on several trade-off issues in partitioning the problem across WAN nodes, and LAN/WAN networks of tightly coupled computing clusters. Time permitting, this talk will conclude with a brief overview of some of Prof. Casavant’s other research areas, including some applications of Bioinformatics to Medical Genetics in disease gene mutation finding.
May 4 - Lilach Hadany, Depts. of Biological Sciences and Mathematics, University of Iowa, "Stress, genetic variation, and their interaction: new evolutionary models".
Abstract: Genetic variation provides the raw material for evolutionary change. In most population genetics models, variation is assumed to be generated at a uniform rate, depending on the genes coding for variation but not on the state of the individual. In this talk I discuss the implications of a new assumption - that genetic variation is generated at higher rates under stress. Using analytical models and computer simulations we found that stress-induced genetic variation can evolve under a wide parameter range, and might help explain the evolution of sex and the mechanisms of complex adaptation. I will discuss theoretical models and experimental evidence.

Spring 2006

May 5 - Dr. Rob Dittmar, Senior Statistician, CitiMortgage, Inc, "An Introduction to the Mathematics of Bond Pricing".
April 28 - Prof. Colleen Catharine Mitchell, UI Dept of Mathematics, "Neural Timing in Highly Convergent Systems".
April 21 - Prof. Debashish Bhattacharya, UI Dept. of Biological Sciences, "Using Comparative Genomics and Phylogenetics to Understand the Origin of Photosynthesis in Eukaryotes".
April 14 - Prof. Chun-Fang Wu, UI Dept. of Biological Sciences, "Temporal Patterns of Neuronal Action Potential Trains Underlying Stereotyped, Fixed Actions and Plastic Behaviors".
April 7 - Prof. Jian Huang, UI Dept. of Statistics and Actuarial Science, "Regularized Estimation with High Dimensional Data".
March 31 - Prof. Matt Schuette, Mathematical Sciences, Georgia Southern University, Statesboro, "A Model of the Transmission of Feline Leukemia Virus Including Vector Spread by the Cat Flea".
March 24 - Prof. Ching-Long Lin, UI Mechanical and Industrial Engineering, "Multiscale Simulation of Air Flow in the CT-based Human Lung Model".
February 24 - Prof. John Logsdon, UI Dept of Biological Sciences, "Phylogenomics of sex: finding meiosis genes and analyzing their evolution".
February 17 - Prof. Kurt Anstreicher, UI Dept of Management Sciences, "An improved algorithm for computing Steiner minimal trees in Euclidean d-space".
February 10 - Prof. Yangbo Ye, UI Dept. of Mathematics, "Computed tomography reconstruction from cone-beam scanning along an arbitary curve".
February 3 - Prof. Herb Hethcote, UI Dept of Mathematics, "History of AMCS and Its Students".
January 27 - Prof. Steve Hendrix, UI Dept of Biological Sciences, "Empirical and Modeling Studies of Dispersal and Diversity in Fragmented Landscapes".

1.23