news and events > talks
Talks, Seminars and Guest Lectures
CSCS Seminars -- Winter 2008
Upcoming seminars
TBD
Past seminars in Fall 2007
"Topological Transitions and Singularities in Fluids: The Life and
Death of a Drop"
April 3, 2008
335 West Hall
1pm
Prof. Sidney Nagel
--Department of Physics, University of Chicago
The exhilarating spray from waves crashing into the shore, the
distressing sound of a faucet leaking in the night, and the
indispensable role of bubbles dissolving gas into the oceans are but a
few examples of the ubiquitous presence and profound importance of drop
formation and splashing in our lives. They are also examples of a
liquid changing its topology. Although part of our common everyday
experience, these transitions are far from understood and reveal
delightful and profound surprises upon careful investigation. For
example in droplet fission, the fluid forms a neck that becomes
vanishingly thin at the point of breakup. This topological transition
is thus accompanied by a dynamic singularity in which physical
properties such as pressure diverge. Singularities of this sort often
organize the overall dynamical evolution of nonlinear systems. I will
first discuss the role of singularities in the breakup of drops. I will
then discuss the fate of the drop when it falls and eventually splashes
against a solid surface.
"The Complexity of Dangerous Dyads --A First Cut: Democracy and Distance"
April 10, 2008
335 West Hall
1pm
Bear Broumoeller
--Political Science, Ohio State University
Quantitative international relations scholars occasionally, and
grudgingly, admit that the nearly 40,000 dyads in the world might not
have the same baseline propensity for conflict: Botswana and Bolivia,
for example, seem quite a bit less likely to fight than India and
Pakistan. A crude attempt to compensate for this fact has been made by
only using data consisting of "politically relevant dyads," but the
attempt is crude indeed: over 25% of militarized disputes take place
between states deemed non-relevant by the usual measure. I attempt to
understand theoretically how both distance and democracy have an impact
on conflict propensity, and from that theoretical understanding I derive
a statistical model and show that it improves the overall fit of some
well-known existing studies of conflict.
"Cellular Evolution in the Origin and Development of Cancer"
March 27, 2008
335 West Hall
1pm
John Pepper
--Ecology & Evolutionary Biology, University of Arizona
Natural selection can act not only among organisms within a species,
but also among cells within a multicellular organism. The resulting
'somatic' cellular evolution poses a serious threat to animals that
undergo continual cell multiplication during their life. Mutant cells
with increased survival and replication arise easily, and tend to out-
compete and replace the less 'fit' cells that remain constrained by
organismal controls on survival and replication. The end result of
this process is uncontrolled cell proliferation, or cancer. Long-lived
organisms are viable only because they have evolved mechanisms to
suppress within-organism cellular evolution. Here we hypothesize that
a specific pattern of cell differentiation, typical of animal tissues,
functions as a primary defense against cancer by blocking cellular
selection. A corollary is that the cancer process begins with the
disruption of these normal cell differentiation patterns. We use an
agent-based computational model to develop this hypothesis and explore
its implications. The results fit a number of observed patterns, and
also generate novel and testable predictions.
"Reasoning at Uncertainty-Insights from the Text of Securities Analysts' Reports and Mathematical Modeling"
March 20, 2008
335 West Hall
1pm
Timo Ehrig
--University of Illinois, Chicago
In new and turbulent markets, like the ".com" stock market in the late 1990ies, the
translation of news, rumors, and data into stock prices is not well defined.
Our model accounts for a multiplicity of possible perspectives and forecasts. We do
believe that there is no "true" prediction model, but analysts, investors, and other
market actors negotiate and debate possible futures that are partly co-created by the
forecasts.
Analysts create frameworks using metaphors and analogies to organize present data and to
derive price targets (Beunza/ Garud 2007).
The analysts dealt with what we dub the "translation problem": How do you translate thing
A that seems similar to thing B in the linguistic domain into the measures of thing B?
The linguistic space that frames a stock at uncertainty spans a logical space, in which
present and future events can be sorted and have logical consequences. We mobilize
results from psychology and linguistics such as Weick´s sensemaking or Lakoff´s work on
metaphors for the development of our economic model. This model formalizes the
exploration of possibility spaces: Different consequences result from different
translations of something into the measures of an analogous known thing.
At heart of our model is a diagramming technique that we use to translate logical
entailments in ordinary language into logical diagrams.
After the talk, there will be time for an experiment with interested participants to try
out the diagramming technique first hand.
"Deciphering Complex Neuronal Processes in C. elegans using novel
Microfluidic Devices"
March 13, 2008
335 West Hall
1pm
Nikos Chronis
--UM Mechanical and Biomedical Engineering
How information is processed in neural networks to generate behavior is
a fundamental question in neuroscience. How, for example, is mechanical
stimulation perceived and converted to an unpleasant feeling or a
repulsive behavior? Among the various animal models that have had their
nervous system studied so far, the nematode C. elegans has been proven
to be an excellent model for bridging the gap between behavior and
neural circuits due to its optically transparent and compact nervous
system as well as due to its genetic conservation with vertebrate
nervous systems. In this talk, I will introduce novel micro-engineered
devices (microfluidics) that can facilitate the understanding of complex
neuronal processes in C. elegans at the network-level (generation of
locomotion) as well as at the single neuron-level (neuronal aging, axon
regeneration).
"Blasting, Packing, and Zipping: Connections between Sequence Alignment, Information Theory, and Data Compression"
February 21, 2008
335 West Hall
1pm
Peter Grassberger
--University of Calgary
Usually, sequence alignment is based on optimization of heuristic scoring schemes. We
show that these scoring schemes can be replaced by mutual information, allowing thus an
objective and inherent comparison of the quality of different alignment algorithms. At
the same time, this provides a link between "zipping-based" clustering schemes (which
have a theoretical foundation in algorithmic information theory) and more conventional
algorithms for generating phylogenetic trees based on
sequence alignment. In particular we show that estimates of the
``normalized universal similarity distance" between mammalian mitochondrial DNA can be
much improved by using global alignment plus
zipping, instead of zipping alone. Finally, we speculate on using ideas
and methods from local sequence alignment for improving general purpose
data compression algorithms.
"The Dynamic Spread of Happiness in a Large Social Network"
February 7, 2008
335 West Hall
1pm
James Fowler
--Associate Professor, Political Science, UC San Diego
The study of happiness is receiving increasing attention in economics,
psychology, neuroscience, and evolutionary biology, and a broad range of
stimuli to human happiness and unhappiness have been explored, including
lottery wins, income, job loss, socioeconomic inequality, divorce,
commuting time, illness, bereavement, and genes. However, these studies
have not addressed a key stimulus to human happiness: the happiness of
others. Past work has shown that people imitate facial expressions , and
that one persons mood might fleetingly determine the mood of others.
However, whether happiness spreads broadly and more permanently across
social networks is unknown. Here, we measure the happiness of 5,019
individuals over 18 years in the Framingham Heart Study Social Network
(FHS-Net) to examine how network characteristics affect a persons
happiness and how happiness spreads between friends, spouses, siblings,
and neighbors. Clusters of happy and unhappy people are visible in the
network, and the effect of one persons happiness appears to extend up to
three degrees of separation (to ones friends friends friends). Moreover,
network characteristics predict which individuals will be happy in the
future, particularly those who are surrounded by many happy people and
those who are highly central in their local networks. We also find that
the spread of happiness decays with both physical distance and time.
Finally, longitudinal statistical models suggest that happiness clusters
result from the spread of happiness and not just a tendency for people
to associate with those who exhibit similar emotional states.
"Robustness in Complex Biological Networks"
January 31, 2008
335 West Hall
1pm
Hans Othmer
--Dept of Mathematics and Digital Technology Center, University of Minnesota
Biological networks that arise in signal transduction, metabolism,
gene control or other cellular functions frequently involve many steps
and many levels of control, but are remarkably reliable in producing
the desired output in response to inputs. In this talk we will
discuss general characteristics such as sensitivity and adaptation
of networks, give several examples that illustrate how robustness is
achieved, and discuss the mathematical techniques that can
contribute to understanding complex networks.
"Physical Limits to Biochemical Signaling"
January 24, 2008
335 West Hall
1pm
Sima Setayeshgar
--Physics, University of Indiana
Biochemical reactions constitute the cell's computing language, often
operating with surprisingly small numbers of molecules. There is renewed
interest in analyzing the impact of noise associated with these small
numbers and the reliability with which cell's crucial tasks can be
carried out. Thirty years ago, Berg and Purcell showed that bacterial
chemotaxis, where a single-celled organism must respond to small changes
in concentration of chemicals outside the cell, is limited directly by
molecule counting noise and that aspects of the bacteria's behavioral
and computational strategies must be chosen to minimize the effects of
this noise. In this talk, I will revisit and generalize their arguments
to estimate the physical limits to signaling processes within the cell
and argue that recent experiments are consistent with performance
approaching these limits. Performance near the limits allowed by the
laws of physics appears as a common theme in biology that cuts across
scales, systems and organisms.
HIV Risk Behavior Fluctuation, Genetic Patterns, and Control: New
Complex Systems Insights and Methods
Thursday, December 13, 2007
335 West Hall
1pm
James Koopman, MD, MPH School of Public Health UM
HIV prevalence is extremely sensitive to fluctuation by individuals
between high risk and lower risk behavior modes. The sensitivity arises
due to a peak of transmissibility during primary HIV infection (PHI).
This sensitivity is a likely explanation for the persistence of rather
constant new infection rates in the US despite improved diagnosis and
treatment. Most model analyses have ignored behavior fluctuations
because of poor data documenting them. All experts agree, however, that
such fluctuations are a reality. We show that such fluctuations
dramatically change the potential population level effects of
controlling transmission at different stages. Furthermore we show that
data such as that gathered in Montreal on HIV sequence relationships
between viruses isolated from individuals diagnosed with PHI or later
stages of infection can help reveal how transmission dynamics are likely
to affect population level effects of stopping transmission at different
stages of infection.
The (Re) Electrification of the Automobile
Thursday, November 29, 2007
335 West Hall
4pm (note time)
Michael Tamor
Executive Technical Leader, Hybrid and Fuel Cell Vehicle Research
Ford Research and Advanced Engineering
Abstract:
It is easily forgotten that a century ago electric vehicles had a
dominant market share over steam and internal combustion automobiles.
While quiet, clean and reliable, the EV was challenged by range and
performance limitations, and eventually done in by the electric
self-starter...one of the first electrified features! Driven in
large part by the quest for improved fuel efficiency and emissions --- and
accelerated by design opportunities and
attractive customer features --- the re-electrification
of motor vehicle is now progressing rapidly. Electrified functions
range from the mundane, such as power steering and engine coolant
pumps, through a spectrum of hybrid powertrain concepts and on to a
resurgent interest in all-electric vehicles. Each incremental function
and additional load increases the demand for efficient and
cost-effective energy storage and is driving new, sometimes divergent,
battery requirements. The hybrid electric vehicles available today are
all what might be called "power hybrids" that have been optimized to
achieve impressive fuel efficiency with minimum battery energy
capacity --- and so minimum battery weight and
cost. Driven by commercialization of HEVs, rapid progress on batteries
and power electronic systems paves the way to two revolutions in the
motor vehicles: First, is the proposal to shift a significant fraction
of propulsion energy from petroleum fuel to electricity by "plugging
in." The "plug-in" HEV is an "energy hybrid" that by definition
requires a much larger battery, and creates a new connection that
allows motor vehicles to exploit the efficiency and diversity of the
electric grid. While not overwhelmingly attractive in the present
economy of still-cheap oil and fossil fuel-generated electricity, the
plug-in hybrid could prove critical to closing the gap between a
limited supply of renewable fuel and ever-growing transportation
needs. Second, with no rigid connections between the various
powertrain elements, the series HEV --- "power" or
"energy" --- enables re-engineering of the motor vehicle including its
shape, structure, materials and manufacturing. In short, the
re-electrification of the automobile is enabled by progress in battery
technology, but itself may enable revolutions in the nature and the
economics of personal transportation.
A Mathematical Model for Vibrio Cholerae Colonization of the Human Intestine
Thursday November 15, 2007
335 West Hall
1:00 pm
Anna Maria Spagnuolo
Department of Mathematics and Statistics, Oakland University.
Anna Spangnuolo's homepage
Abstract:
Vibrio cholerae is a strict human pathogen that causes pandemic cholera.
It is an old-word pathogen that has re-emerged as a new threat since the early
1990s. V. cholerae colonizes the upper, small intestine where it produces a toxin
that leads to the watery diarrhea, characterizing the disease. Colonization dy-
namics of the bacteria are largely unknown. Although a large initial infectious
dose is required for infection, data suggests that only a smaller sub-population
colonizes a portion of the small bowel leading to disease. There are many barri-
ers to colonization in the intestines including peristalsis, fluid wash-out, viscosity
of the mucus layer, and pH. It is of interest to identify the mechanisms that al-
low this sub-population of bacteria to survive and colonize the intestines when
faced with these barriers. In this talk, I will elaborate on the dynamics of V.
cholerae infection by describing a mathematical model that governs the colo-
nization process for the bacterial dynamics. The results indicate that both host
and bacterial factors contribute to colonization. Host factors include intestinal
diffusion and convection rates while bacterial factors include adherence, motility
and growth rates. This model can be used to test therapeutic strategies against
V. cholera.
An Example Using Cultural Algorithms
Thursday November 8, 2007
335 West Hall
1:00 pm
Robert G. Reynolds
Department of Computer Science, Wayne State University
Adjunct Associate Research Scientist, Museum of Anthropology University
of Michigan-Ann Arbor
Abstract:
Cultural Algorithms are an extension of Genetic Algorithms that employ
a basic set of knowledge sources, each related to knowledge observed in
various animal species. These knowledge sources are then combined to
direct the decisions of the individual agents in solving optimization
problems. Here an algorithm based upon an analogy to the marginal value
theorem in foraging theory is developed to guide the integration of
these different knowledge sources to direct the agent population within
the Cultural Algorithm framework. The Algorithm is applied to finding
the optimum in a dynamic environment composed of mobile resource cones.
It is demonstrated that certain phases of problem solving can emerge
along with related individual roles during the solution process.
Discovering the Molecular Basis of Biological Oscillations
Monday, October 29th
335 West Hall
noon
Danny Roger
UM Math Department
Abstract:
Biological clocks time many biological processes that must occur at
specific times of the day. The biological basis for these clocks is
genetic feedback loops that can sometimes produce oscillations in the
concentration of proteins. While most genetic feedback loops contain
properties which might make them candidates for clocks (e.g. feedback
and delay) only a few actually oscillate. I will describe the
mathematical reasons why sustained oscillations in genetic systems are
very difficult to achieve and several biological strategies that can be
used to overcome these constraints.
NONLINEAR DYNAMICS OF THE 3D PENDULUM
Thursday, September 13, 2007
335 West Hall
1pm (NOTE TIME)
N. Harris McClamroch
Department of Aerospace Engineering, U of M
A 3D pendulum consists of a rigid body, supported at a fixed
pivot, with three rotational degrees of freedom. The pendulum is acted
on by gravity. Symmetry assumptions are shown to lead to the planar 1D
pendulum and to the spherical 2D pendulum models as special cases. The
case where the rigid body is asymmetric and the center of mass
is distinct from the pivot leads to the concept of the 3D pendulum.
Several 3D pendulum models are introduced and used to study important
features of the nonlinear dynamics: conserved quantities, equilibria,
invariant manifolds, local dynamics near equilibria and invariant
manifolds, and the presence of chaotic motions. These results
demonstrate the rich and complex dynamics of the 3D pendulum. Graphics
and animations will be presented to illustrate the results.
|
