Calendar

COLLOQUIUM OF THE COMPUTATIONAL MATERIALS SCIENCE CENTER
AND THE DEPARTMENT OF COMPUTATIONAL AND DATA SCIENCES
(CSI 898-Sec 001)

GeoSocial Media Analysis: Challenges and Opportunities

Arie Croitoru
The Center for Geospatial Intelligence
Department of Geography and Geoinformation Science
George Mason University
Fairfax, VA

Over the last few years, social media (e.g. Twitter, Flickr, YouTube, etc.) have become an integral part of the modern information and communication landscape. Through social media, individuals, groups, organizations, and even states can now acquire, probe, and deliver information, as well as shape and reshape public opinion. At the same time, social media content is increasingly related to physical geographical locations. Fueled by advances in of Web 2.0, mobile computing, and spatially-aware technologies (i.e. GPS enabled smartphones), social media can provide a unique opportunity to observe and study the flow of information in both cyber and physical spaces. Employing a geographically data-driven analysis approach allows us not only to track how information spreads, but also to derive information about real-world events and happenings. We call this geospatially-driven approach “GeoSocial Media Analysis.”   In this presentation I will review some of the key trends and developments that has enabled the rise of geosocial media, and shaped their characteristics. I will then move on to describe examples of various geosocial analysis methods in the context of different application areas (e.g. natural disasters, protests, international relations). Looking ahead, I will conclude the presentation with some of the emerging challenges and needs in the analysis of geosocial media data.

September 21, 2015, 4:30 pm
3301 Exploratory Hall
Fairfax Campus

Refreshments will be served at 4:15 PM.

Find the schedule at http://www.cmasc.gmu.edu/seminars.htm

COLLOQUIUM OF THE COMPUTATIONAL MATERIALS SCIENCE CENTER
AND THE DEPARTMENT OF COMPUTATIONAL AND DATA SCIENCES
(CSI 898-Sec 001)

Exotic Superconductors

Ahmet Keles
Department of Physics and Astronomy
George Mason University
Fairfax, VA

Superconductivity has been one of the main problems in condensed matter physics a hundred years after its discovery. Many results of this intensive research made it to textbooks hinting at its fundamental importance besides promising applications. In the last few decades, new superconducting materials were found that were beyond the realm of conventional phonon based superconductors described by Bardeen-Cooper-Schriffer (BCS) theory. This new class of superconductors found a wide ground in particle physics due the presence of exotic symmetries that can be described by models akin to the Lorentz invariant field theories.

In this talk, I will be focusing on a recent member of this family of exotic superconductors Weyl superconductors and related Majorana fermions and discuss possible experimental realizations of these materials.

September 28, 2015, 4:30 pm
Exploratory Hall, Room 3301
Fairfax Campus

Refreshments will be served at 4:15 PM.

Find the schedule at http://www.cmasc.gmu.edu/seminars.htm

COLLOQUIUM OF THE COMPUTATIONAL MATERIALS SCIENCE CENTER
AND THE DEPARTMENT OF COMPUTATIONAL AND DATA SCIENCES
(CSI 898-Sec 001)

Mining Big and Complex Data: Challenges and Solutions

Carlotta Dominicone
Department of Computer Science
George Mason University
Fairfax, VA

The need for mining massive and complex data has become paramount in areas like education, web mining, social network analysis, security, and a variety of scientific pursuits. However many traditional supervised and unsupervised learning algorithms break down when applied in big data scenarios: this is known as the “big data problem”. Among other concerns, big data presents serious scalability difficulties for these algorithms.

In this talk I will discuss the big data problem and solutions in the context of complex data I have worked on in recent years, namely, social networks, text, and educational data. I will also present a new method for distributed machine learning which directly tackles key problems posing challenges to successful and scalable mining of big data. Our method is scalable, general-purpose with regard to the machine learning algorithm, and easily adaptable to a variety of heterogeneous grid or cloud computing scenarios.

October 5, 2015 – 4:30 p.m.
Exploratory Hall, Room 3301
Fairfax Campus

Refreshments will be served at 4:15 PM.

Find the schedule at http://www.cmasc.gmu.edu/seminars.htm

COLLOQUIUM OF THE COMPUTATIONAL MATERIALS SCIENCE CENTER
AND THE DEPARTMENT OF COMPUTATIONAL AND DATA SCIENCES
(CSI 898-Sec 001)

Co-evolutionary Algorithms: A Useful Computational Abstraction?

Kenneth De Jong
Department of Computer Science and Office of Research Computing
George Mason University
Fairfax, VA

October 19, 2015, 4:30 p.m.
Exploratory Hall, Room 3301

Interest in co‐evolutionary algorithms was triggered in part with Hillis’ 1990 paper describing his success in using one to evolve sorting networks. Since then there have been heightened expectations for using this nature‐inspired technique to improve on the range and power of evolutionary algorithms for solving difficult computation problems. However, after more than two decades of exploring this promise, the results have been somewhat mixed.   In this talk I summarize the progress made and the lessons learned with a goal of understanding how they are best used and identify a variety of interesting open issues that need to be explored in order to make further progress in this area.

Refreshments will be served at 4:15 PM.

Find the schedule at http://www.cmasc.gmu.edu/seminars.htm

COLLOQUIUM OF THE COMPUTATIONAL MATERIALS SCIENCE CENTER
AND THE DEPARTMENT OF COMPUTATIONAL & DATA SCIENCES
(CSI 898-Sec 001)

Molecular dynamics simulations of peptides relevant to Alzheimer’s disease

Christopher Lockhart
School of Systems Biology
George Mason University
Manassas, VA

The progression of Alzheimer’s disease has been linked to the aggregation of Aβ peptides. Current theories suggest that Aβ interactions with cellular membranes results in its neurodegenerative effect, although little is known of the molecular mechanism. To probe this mechanism, we performed replica-exchange molecular dynamics simulations of Aβ peptides binding to a model membrane represented by a lipid bilayer.

In this talk I will describe our investigation of the conformational change of Aβ peptides induced by the bilayer, the Aβ binding mechanism, and the impact of Aβ on bilayer structure. Our results find a correlation between the Aβ cytotoxicity and the aggregation of Aβ peptides.

October 26, 2015 – 4:30 p.m.
Exploratory Hall, Room 3301
Fairfax Campus

Refreshments will be served at 4:15 PM.

Find the schedule at http://www.cmasc.gmu.edu/seminars.htm

COLLOQUIUM OF THE COMPUTATIONAL MATERIALS SCIENCE CENTER
AND THE DEPARTMENT OF COMPUTATIONAL AND DATA SCIENCES
(CSI 898-Sec 001)

NEW TRANSITORY PHASES OF SILICA UNDER HIGH PRESSURE

Qingyang Hu
Geophysical Laboratory
Carnegie Institution of Washington
Washington, D.C.

Silica is one of the most-abundant natural compounds and a major component of the Earth’s crust and mantle. Its various high-pressure forms make it an often-used study subject for scientists interested in the transition between different chemical phases under extreme conditions, such as those mimicking the deep Earth. Compressing single crystal coesite SiO2 under hydrostatic pressures of 26~53 gigapascal at room temperature, we discover a new polymorphic phase transition mechanism of coesite to post-stishovite, by means of single-crystal synchrotron x-ray diffraction experiment and first-principles computational modeling.

The transition features the formation of multiple previously unknown triclinic phases of SiO2 on the transition pathway as structural intermediates. Coexistence of the low-symmetry phases results in extensive splitting of the original coesite X-ray diffraction peaks that appear as dramatic peak broadening and weakening, resembling an amorphous material. This work provides new insights into the structural transition of SiO2 crystal under high pressures, and clarifies the issue of the pressure-induced amorphization of coesite, which has often been cited as an archetypal example of the amorphization phenomena in general.

November 2, 2015 – 4:30 p.m.
Exploratory Hall, Room 3301
Fairfax Campus

Refreshments will be served at 4:15 PM.

Find the schedule at http://www.cmasc.gmu.edu/seminars.htm

COLLOQUIUM OF THE COMPUTATIONAL MATERIALS SCIENCE CENTER
AND THE DEPARTMENT OF COMPUTATIONAL AND DATA SCIENCES
(CSI 898-Sec 001)

Auger Recombination And Its Suppression In Nanoscale Semiconductors

Roman Vaxenburg
Computational Material Science Center
George Mason University, Fairfax, VA
and Naval Research Laboratory, Washington, D.C.

November 9, 2015, 4:30 p.m.
Exploratory Hall, Room 3301

Nonradiative Auger recombination is an energy dissipation mechanism in which the recombination energy of an electron-hole pair, instead of being emitted as a photon, is transferred to another charge carrier and is eventually lost as heat. If efficient enough, the Auger process can considerably reduce the radiative efficiency of semiconductor materials, therefore complicating their application as light emitters. As a consequence of quantum confinement, in nanoscale semiconductor systems such as quantum wells or quantum dots, the Auger recombination is greatly enhanced and is able to interfere with the radiative recombination. Here we present a theoretical study of the nonradiative Auger recombination processes in III-V quantum well light-emitting diodes (LEDs). The performance of these LEDs is known to steadily decrease with increasing driving current and the origin of this efficiency droop phenomenon is still debatable. Our calculations show that the Auger recombination can be efficient enough in these materials and that it can be the source of the efficiency droop. Further, we propose a strategy to suppress the confinement-enhanced Auger recombination by controllable modification of the confining potential shape. We apply this approach to both polar and nonpolar variants of the III-V quantum well LEDs and we demonstrate an appreciable suppression of the Auger recombination and consequent improvement of light-generation efficiency in these materials.

Refreshments will be served at 4:15 PM.

———————————————————————-

Find the schedule at http://www.cmasc.gmu.edu/seminars.htm

COLLOQUIUM OF THE COMPUTATIONAL MATERIALS SCIENCE CENTER AND
THE DEPARTMENT OF COMPUTATIONAL AND DATA SCIENCES
(CSI 898-Sec 001)

Insights into Cardiac Arrhythmia using a Multiscale Systems Biology Approach

Saleet Jafri
Department of Molecular Neuroscience
Krasnow Institute for Advance Studies
George Mason University
Fairfax, VA

November 16, 2015 – 4:30 pm
Exploratory Hall, Room 3301
Fairfax Campus

Heart disease is the leading cause of death in the developed world and is being recognized as and increasing problem in the the developing world. A fatal cardiac arrhythmia is often the underlying cause of death. One class of arrhythmia rises from defects in calcium dynamics in the heart. Calcium plays a major role on heart function linking the electrical depolarization of the heart with contraction. We use an integrative multiscale model to explore how defects on the level of molecular function can lead to arrhythmia in the heart.

These simulations cross many orders of magnitude on the spatial and temporal scales to provide insight into mechanisms of cardiac arrhythmia. To make this problem tractable we have employed novel computational algorithms and leverage modern GPU computing techniques. The insights gained by these studies provide advanced our understanding of arrhythmia such as how an arrhythmia is initiated, how genetic defects can contribute to arrhythmia, and how do changes to structure and gene expression during disease increase the likelihood of arrhythmia.

Refreshments will be served at 4:15 PM.

———————————————————————-

Find the schedule at http://www.cmasc.gmu.edu/seminars.htm

COLLOQUIUM OF THE COMPUTATIONAL MATERIALS SCIENCE CENTER
AND THE DEPARTMENT OF COMPUTATIONAL & DATA SCIENCES
(CSI 898-Sec 001)

DETERMINING OPTIMAL POLICE PATROL AREAS WITH MAXIMAL COVERING AND BACKUP COVERING LOCATION MODELS

Kevin Curtin
Department of Geography and Geoinformation Science and,
Department of Computational and Data Sciences
George Mason University,
Fairfax, VA

In this talk I will present a new method for determining efficient spatial distributions of police patrols in a metropolitan region. A mathematical model called the police patrol area covering (PPAC) model is formulated and solved. The method uses inputs of police geography (divisions, sectors, beats, and response areas) within a geographic information systems (GIS) framework, analyzes that data through an optimal covering model formulation, and provides alternative optimal solutions for presentation to decision makers. The goals of this research are to increase the level of police service in the face of variation in both the demand for police services and the availability of police resources. Examples of the inputs from – and outputs to – GIS are provided through a pilot study of the city of Dallas, Texas.

November 30, 2015 – 4:30 p.m.
Exploratory Hall, Room 3301
Fairfax Campus

Refreshments will be served at 4:15 PM.

Find the schedule at http://www.cmasc.gmu.edu/seminars.htm

COLLOQUIUM OF THE COMPUTATIONAL MATERIALS SCIENCE CENTER AND
THE DEPARTMENT OF COMPUTATIONAL AND DATA SCIENCES
(CSI 898-Sec 001)

Neutron Scattering from Nanostructured Oxide Materials                      

Krishnamurthy V. Vemuru
Department of Physics and Astronomy
George Mason University
Fairfax, VA

February 8, 2016 – 4:30 pm
Exploratory Hall, Room 3301
Fairfax Campus

Neutrons can penetrate deeply into materials. They can sense the position of atomic nuclei through strong interaction and magnetic moments through dipole-dipole interaction. The scattering of neutrons by both these mechanisms is well understood, so theories need not be concerned about the modeling or simulating the scattering process. Instead, the investigations can focus on the computation and experimental determination of the intrinsic positions of nuclei and spin-correlation functions. This enables neutron scattering as a unique tool to determine crystal structures, nanostructures, spin dynamics and phonon dispersions in solids.   In this talk, I will focus on the applications of small angle neutron scattering to determine the orientational ordering of nanostructures in MgO nanorods, TiO2 nanowires and gamma-Fe2O3 nanoparticle dispersions.   I will also discuss how small neutron angle scattering from nanostructures is modeled and calculated.

Refreshments will be served at 4:15 PM.

———————————————————————-

Find the schedule at http://www.cmasc.gmu.edu/seminars.htm