Search Google Appliance


Seminar Archive 2007

 

 


Date:
November 30, 2007

Location:
Harder House, Room 104

Time:
1-2 pm

Presenter(s):
Olgay Cangur

Title:
Modeling Subprime Delinquency, Termination and Loss

Abstract:
The proposed research focuses on several topics relevant to modeling residential mortgages while each topic is described separately, they will, in fact, be done in parallel and each influence the other. The first topic provides a framework for modeling prepayments and defaults that represents an enhancement over previous studies. A total of nine loan payment statuses will be used (current, thirty days late, sixty days late, ninety days late, early foreclosure, late foreclosure, real estate owned, paid off, and terminated with loss). This framework will be compared to the previous framework discussed in the literature that uses seven statuses.

The second topic will investigate the effect of a servicer's loan workout and their loss mitigation efforts on the finalized loan loss. Effect of the key variables related to loan workout and loss mitigation efforts will be tested for significance. A stand-alone loss model will be built to predict expected losses for a horizon of 6, 12, and 18 months incorporating these findings. The result will be compared to the existing loss models found in the literature.

The third topic will apply reconstructability analysis (RA) to modeling residential mortgage data in order to find new and interesting models. Many statistical methods are unable to reflect non-linearities and significant high-level interactions. RA is capable of doing both. The new modeling framework and the loss model, mentioned as the first and second topics respectively, will utilize the findings of the RA research. In order to prove the effectiveness of RA, results will be tested with actual data and compared with similar statistical models.

Bio(s):
Olgay has been a Systems Science student since 2002. He is currently working on his dissertation. He is also a full time employee of Merrill Lynch. His areas of interest are forecasting mortgage delinquencies.

File attached or Link to Recording?
None



Date:
November 16, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Dr. Niles Lehman

Title:
Self-replication and autocatalysis: How can chemicals come alive?

Abstract:
We are investigating the chemical origins of life on the Earth. According to practical definitions of "life" a collection of chemicals can only be alive if they can self-replicate, and this requires a chemical property known as autocatalysis. We have engineered a system of RNA molecules that we believe has this property, in that a collection of short oligomeric RNAs can spontaneously self-assemble into a self-replicating catalytic RNA. In this talk I will discuss the chemical nature of life itself and how our system may shed some insight into the abiotic-biotic transition on the Earth 4 billion years ago.

Bio(s):
Dr. Lehman is a Professor in the Department of Chemistry at Portland State University. He received his Ph.D. in Biology at UCLA, and his M.A. in Comparative Biochemistry at UC Berkeley. His research projects include quantitative studies of in vitro evolution of catalytic RNA, and computer parameterization, information-theoretical analysis, and modeling of the origins of life and of RNA evolution in vitro. He is an associate editor of the Journal of Molecular Evolution, an editorial board member of Astrobiology, and a faculty member of the Center for Life in Extreme Environments.

File attached or Link to Recording?
None



Date:
November 9, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Cecily Froemke

Title:
National Material Capabilities Prediction of Conflict: A Non-Linear Approach Using Reconstructability Analysis.

Abstract:
This paper applies reconstructability analysis (RA), an approach developed in the systems science community and a tool of discrete multivariate modeling, to a selection of variables contained in data sets from the Correlates of War Project. The goal of this paper is to contribute the methods of reconstructability analysis to the field of international conflict. Specifically, I will look at the following questions: To what extent does a nation's material capabilities affect whether or not the nation becomes engaged in an international conflict? Additionally, how well do these material capabilities do in predicting the outcomes of international conflict?

Bio(s):
I am currently a graduate student in the Systems Science program and studying for comprehensive exams in the spring. I have an interest in all the elements and relations of Systems Science but am especially interested in information theory, from multiple perspectives at multiple scales. Discrete Multivariate Modeling is one of my more technical forays in this arena.

As for that thing I do from 8-5: I work as an Application Specialist with a local software company that participates in outcomes data management for cardiovascular, thoracic, transplant, orthopedic and vascular surgery.

I truly enjoy dialogue with folks from all walks of life, inside and outside of academia. Unfortunately, work has kept me from the wonderful happenings at Harder House lately, but I am looking forward to talking with you all on Friday!

File attached or Link to Recording?
None



Date:
November 2, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Michael S. Johnson

Title:
Report back from M2007 Data Mining Conference

Abstract:
The annual MYYYY data mining conference brings together academic researchers, practitioners from a range of industries, and hardware/software vendors for presentations and discussions about data mining methods and applications. The conference is held in Las Vegas, home to some of the world's most experienced and sophisticated data mining practitioners (employed, for the most part, by the gaming industry). I'll provide an overview of this year's conference and will seed the discussion with some observations based on the presentations I attended.

Bio(s):
Michael S. Johnson (SYSC Ph.D. 2005) is the Director of the Utility for Care Data Analysis (UCDA), an analytical department within the Program Office of the Kaiser Permanente health care organization (KP). The UCDA was created in 2005 to improve the quality of care and service for KP's 8.7 million members by applying advanced analytical methods to the data accumulating in KP's electronic medical record system.

File attached or Link to Recording?
None



Date:
October 26, 2007

Location:
Harder House, Room 104

Time:
1-2 pm

Presenter(s):
Scott Mist

Title:
Can questionnaires be used to predict Traditional Chinese Medicine Diagnosis?

Abstract:
Astonishingly, most Traditional Chinese Medicine (CM) research in the West proceeds without CM diagnoses, which raises serious questions. One way to improve the feasibility of incorporating CM diagnosis would be to prescreen participants using questionnaires. Consequently, Mr. Mist used baseline questionnaires to predict CM diagnosis in 195 participants of a temporomandibular joint disorder study.

Two methods, logistic regression (LR) and reconstructability analysis (RA), were used in conjunction to test Hypothesis 1. Models were created that predicted CM diagnosis from pre-treatment questionnaires. LR models were prepared to predict the diagnosis for each subject using direct effects only. Then variable-based and state-based RA techniques were used to select potentially important interaction terms. These terms were then introduced into the original LR model and assessed for clinical relevance, model simplification, and improved diagnosis prediction. Scott will present the methods and results of these efforts.

Bio(s):
Scott Mist is an acupuncturist and a Traditional Chinese Medicine researcher for the Oregon College of Oriental Medicine. Mr. Mist has worked on research funded by the National Institutes of Health for the last 8 years and is currently the project director for a multisite trial with the University of Arizona. Scott completed a post-graduate research fellowship through the NIH, was granted a NIH Loan Repayment Grant, and is currently completing his PhD in the Systems Science program.

File attached or Link to Recording?
None



Date:
October 19, 2007

Location:
Harder House, Room 104

Time:
1-2 pm

Presenter(s):
Wayne Wakeland

Title:
Effectiveness of a Web-based Training on the Carbon Content of Food

Abstract:
A project is underway to determine the degree to which people's knowledge and attitudes about how their food choices impact the environment can be influenced by participation in a short interactive web-based training. The training utilizes a tool called CarbonScope that was developed recently by Kumar Venkat. The training itself and the pre-test post-test questionnaires were developed by Lindsay Sears, a graduate student in Psychology. The seminar will give some background and then discuss the tool, the training, and the assessment aspects of the study.

Bio(s):
Wayne is a core Systems Science faculty member with a wide variety of research interests, including topics related to computer simulation and sustainability.

File attached or Link to Recording?
None



Date:
October 12, 2007

Location:
Harder House, Room 104

Time:
1-2 pm

Presenter(s):
Joe Fusion

Title:
Environmental Sensitivity and the Evolution of Altruism

Abstract:
We have seen several models of the evolution of altruism involving the Prisoner's Dilemma. This variation takes into account the affects of environmental stresses on the benefits of altruism. I used an agent-based/cellular model, and explored conditions such as hostile vs. friendly environments, and periodic bottlenecks. My current results will be presented, followed by a discussion of future directions.

Bio(s):
Joe Fusion is Ph.D. student in the Systems Science program at PSU. His research interests include artificial life, evolution, and modeling systems. His other interests include everything else.

File attached or Link to Recording?
None



Date:
October 5, 2007

Location:
Harder House, Room 104

Time:
1-2 pm

Presenter(s):
Lars Holmstrom

Title:
Assessing Wind Farm Avian Collision Risk: A Model Based Approach

Abstract:
Wind power is increasingly becoming recognized as a clean and viable renewable energy resource. Despite the fact that conversion to wind power can significantly offset greenhouse gas production on a global scale, there are still numerous environmental impacts to be considered. One of the primary impacts being researched is the effect of wind farms on bird and bat populations, both due to habitat disruption and direct collisions with the wind turbines themselves. This presentation will discuss a model based approach for estimating the mortality risk to birds and bats as a result of these collisions before the construction of a proposed wind farm is even begun. This information is a key step in assessing the environmental impact of these large scale installations and can influence whether construction is ever initiated.

Bio(s):
Lars Holmstrom is a PhD student in the System Science Department at Portland State University. While focusing on statistical, state based modeling and estimation techniques for his dissertation, this presentation reports the results of a summer project performed for a private environmental survey company.

File attached or Link to Recording?
None



Date:
September 28, 2007

Location:
Harder House, Room 104

Time:
1-2 pm

Presenter(s):
None

Title:
Meet-n-Greet

Abstract:
For the first Friday of each term, the Seminar is usually a meet-and-greet type of get-together. For this Friday's Seminar, we will have a welcoming party for all Systems Science students. Refreshments will be served.

Bio(s):
None

File attached or Link to Recording?
None



Date:
June 1, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
None

Title:
System Science Round Table Discussion

Abstract:
This week, we will have our year-end round table discussion with students and faculty to address all things that are Systems Science (well, all things we have time for). This includes questions/gripes/compliments/discussions about courses, departmental direction, web presence, student lounge, etc. This is a great opportunity to touch base with each other and to share ideas. Hope to see you there!

Bio(s):
None

File attached or Link to Recording?
None



Date:
May 25, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Casey Quinlan

Title:
Cardiac protective signaling mechanisms and pathways to mitochondria

Abstract:
Cardioprotection is an endogenous phenomenon whereby the heart protects itself from a myocardial infarction, or heart attack. Cardioprotective drugs mimic this process by triggering cellular signaling events that target the mitochondria. The mechanistic delivery of the signal to mitochondria is an area of some debate. I will present data to support the hypothesis that administration of cardioprotective drugs induces assembly of a plasma membrane signaling platform. Subsequently, this platform interacts with mitochondria and opens the mitochondrial ATP-sensitive K+ channel.

Bio(s):
Casey Quinlan is a fourth year Ph.D. student in the biology department at Portland State University. Her work focuses on mitochondrial physiology and intracellular signaling pathways. She received her undergraduate degree in botany and did a brief foray studying ornithology before she realized that the only hope for civilization lay in mitochondrial bioenergetics. She has since worked tirelessly, with minimal compensation, to bring mitochondria to the people. She also enjoys spaghetti and the song "Dancing Queen" by the Swedish recording artists ABBA.

File attached or Link to Recording?
None



Date:
May 18, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Byrne Lovell

Title:
Reflections on Systems Science and my career as an analyst at the Bonneville Power Administration.

Abstract:
I'm not in academia - I don't do research, and don't have a research specialty to talk about. I work in a fairly large, technically oriented bureaucracy on a variety of analytical tasks that generally can be described as quantitative decision support. I have had a successful career, and am now one of BPA's most senior technical experts. I will talk about how some of the themes from my first years in the Systems Science Ph.D. program relate to my success as an analyst here.

Bio(s):
BA, Math, Pomona College, 1974
MS, Counseling, Oregon, 1980
Ph.D., Systems Science (Uncertainty), PSU, 1995
I started working at BPA in the summer after my first year in the Systems Science Ph.D. Program. I did programming and modeling for the first few years, working with Monte Carlo production cost models and wrote a small, fast model that estimated the market value of BPA's surplus hydro energy. I spent several years in the Office of Finance, and then moved to BPA's Strategic Planning group for five years. I am now in the Chief Risk Officer's group. My particular field of expertise is the financial uncertainty BPA faces due to the highly variable annual supply of water, and various risk mitigation methods we can use in our rates to ensure that we can safely tolerate a very dry year when our power sales revenues are much lower than average.

File attached or Link to Recording?
None



Date:
May 11, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Shari Matzner

Title:
Model-Based Information Extraction from Synthetic Aperture Radar Data

Abstract:
Synthetic Aperture Radar (SAR) is a remote sensing technology that is capable of imaging large areas at high resolution, and can operate at any time, day or night, and in cloudy conditions unlike optical and infrared sensors. SAR has been used primarily for remote monitoring of the natural environment, where the general or average characteristics of an area are of interest. Examples include land use and land cover studies, monitoring tree density in forested areas and estimating soil moisture content.

Conventional processing of the recorded SAR signal produces a two-dimensional image that is an estimate of the surface reflectivity. However, the nature of a SAR reflectivity image is very different from what we are used to seeing with our own optical sensors, our eyes. This makes extracting information about individual structures in the scene challenging. To extract this type of information requires a better understanding of the complicated electromagnetic scattering produced by structures and the impact of the synthetic image processing on these signatures.

This presentation will discuss how a physics-based model of a building can be used in conjunction with a model of the SAR sensor to formulate a building signature, which can then be used in the SAR signal processing to extract information about buildings present in the scene.

Bio(s):
Ms. Matzner is a Systems Science Ph.D. student currently working in the Northwest Electromagnetic and Acoustic Research (NEAR) Lab here at PSU.

File attached or Link to Recording?
None



Date:
April 20, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Josef Lotz

Title:
The development of a remote sensor system for monitoring coral reef morphology and fish abundance

Abstract:
Current methods of coral reef morphology and fish abundance estimation are mature and well known among Fishery Sciences/Engineering researchers. They give highly precise estimations but are dependent on complex technology. Equipment costs have made basic estimation unobtainable to countries and researchers with small budgets. A Matlab toolbox, named EchoMap, is being developed at the NW Electromagnetic and Acoustics Research Lab (PSU) that produces morphology and abundance estimations using a low cost, single-beam echosounder by applying principles of acoustics and signal processing.

Bio(s):
Josef Lotz is a Master's of Science student in the Electrical and Computer Engineering department at PSU. He received his B.S. degree in Electrical and Computer Engineering from Portland State University in 2005. He was a member of the NW Computational Intelligence Laboratory in 2005-2006 and is currently a member of the NW Electromagnetic and Acoustics Laboratory.

File attached or Link to Recording?
None



Date:
April 13, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
James McNames

Title:
Tracking Physiologic Rhythms

Abstract:
Physiologic signals are monitored in many medical applications for a variety of purposes such as diagnosis, prognosis, and locating pathologic tissue. Many of these signals contain rhythms and oscillations due to natural mechanisms, such as the respiratory and cardiac cycles, and pathologic mechanisms such as tremor in Parkinson's disease. Oscillations also occur in natural vocalizations such as speech, animal calls, and echo location. A wide variety of methods have been employed to study these signals ranging from techniques based on chaos theory, time series analysis, hidden Markov models, and detection theory. Members of the Biomedical Signal Processing Laboratory at Portland State University have recently started investigating a new approach based on nonlinear state space models and sequential Monte Carlo methods. During this talk Dr. McNames will define the problem, show several examples of signals from many different domains, summarize the applications, and show preliminary results of our new approach to this problem.

Bio(s):
James McNames received a B.S. degree in electrical engineering from California Polytechnic State University, San Luis Obispo, CA, in 1992. He received M.S. and Ph.D. degrees in electrical engineering from Stanford University, Stanford, CA, in 1995 and 1999, respectively.

He has been with the Electrical and Computer Engineering Department at Portland State University, Portland, OR since 1999, where he is currently an Associate Professor. He has published over 100 peer-reviewed journal and conference papers. His primary research interest is statistical signal processing with applications to biomedical engineering.

He founded the Biomedical Signal Processing (BSP) Laboratory (bsp.pdx.edu) in fall 2000. The mission of the BSP Laboratory is to advance the art and science of extracting clinically significant information from physiologic signals. Members of the BSP Laboratory primarily focus on clinical projects in which the extracted information can help physicians or medical devices make better critical decisions and improve patient outcome.

File attached or Link to Recording?
None



Date:
March 16, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Martin Jetton

Title:
Agent Based Simulation of Sales Floor Activity and Store Sales

Abstract:
In previous modeling of individual sales performance I have found that 'good' sales people can be identified during hiring using a personality trait assessment test. These better sales people perform around 3% better ($/hour) than individuals scoring lower on the personality assessment test percentile score. Using agent based simulation I model two kinds of sales people; those who score well (call them Green sales people) on a sales performance personality assessment and those who did not (call them not-Green sales people). The purpose of this study is to show that store sales variability (the probability that someone will buy at the given store and the amount they will purchase) masks the underlying value of these 'green' sales people if the sales people to customer ratio is low. I show the impact of percentages of green individuals in the store and total store sales relative to customer flow.

One of the reasons I'm using agent based simulation to address this sales person model is the fact that the complexity of human interactions can be tested without impacting the real world. There are a lot of fluffy thoughts around scheduling and testing the parameters of this situation will create a good learning environment to build upon. Human resource management is full of touchy feely individuals and situations that cannot be easily tested in the real world. ABS will allow me to set up a world that could be used to test or baseline rules of interaction on the sales floor relative to customer and sales person characteristics.

In attempting to tackle the chaotic nature of the sales floor in a retail environment, I've explored queuing based models using service models for a structured mathematical approach. I found integrating in other non-queuing rules difficult, if not impossible. I've encountered difficulties in the description and education of clients in the use of modeling environments such as systems dynamics and discrete event simulations, while I find the visual nature of ABS through NetLogo to be perfect to talk to and involve non-technical people in the analytical effort.

Bio(s):
Martin is a Practice Manager, Modeling and Analytics, Talent Management Division, Kronos Inc in Beaverton Oregon. He is enrolled in the Core PhD Systems Science program in his second year of class work. He has a Masters of Science in Operations Research / Applied Statistics, Oregon State University, 1989 and a Bachelor of Science in Mathematical Sciences from OSU, 1985.

Martin has around 17 years of experience ranging from Marketing Research, Product Marketing/Sales Performance analysis, Expert System development, Supply Chain Logistics and Human Capital management analytics.

File attached or Link to Recording?
None



Date:
March 9, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Dan Iancu

Title:
Non-linear interaction between neurons underlie sensory processing in electric fish

Abstract:
The mormyrid electric fish displays extremely fine temporal resolution as measured in behavioral studies. The first stage of sensory processing, the sensory afferents, also display very precise temporal responses to electrical stimulation. The afferents are connected through gap junctions to the granular cells, which in addition receive a corollary discharge signal at the time of the electric pulse. The granular cells have a relatively long time constant, seemingly at odds with their presumed role as coincidence detectors. We use experimental data to build a compartmental model that investigates the mechanism by which the relative timing of the two inputs to the granular cells determines the effect of the afferent spike through the electrical synapse.

Bio(s):
Ovidiu Dan Iancu is currently a graduate student in the Biomedical Engineering PhD program (Oregon Health Sciences University) in Portland, Oregon. He completed his MS degree in Mathematics at Oregon State University (1999). For the last three years he has worked in the Roberts lab, where he studies the biophysical substrates of adaptive sensory processing. The model systems investigated in the Roberts lab include electrosensory fish and the neuronal substrates of song processing in zebra finches.

File attached or Link to Recording?
None



Date:
March 2, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Andrew Toland

Title:
Temporal pattern classification using a biologically inspired coupled oscillator system

Abstract:
Drawing inspiration from one of nature's great pattern classifiers, the olfactory system, I demonstrate that a randomly instantiated coupled oscillator system can be used to improve the classification of time-series signals. The time series used to illustrate the principal are taken from sampled quadruped robot joint positions as the robot walks over surfaces of different characterists using a variety of gaits.

Bio(s):
Andrew Toland is a graduate student in the Systems Science Ph.D. Program. He has a background in physics and biology, and has done some work in image and signal processing. He is currently associated with the NW Computational Intelligence Laboratory, where he is exploring ideas such as the one described in the abstract.

File attached or Link to Recording?
None



Date:
February 23, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
J. Alan Yeakley
Environmental Science
Portland State University

Title:
Nutrient Cycling in Forested Watershed Systems: Responses to Disturbance

Abstract:
I will discuss watershed ecosystem analysis with respect to both hydrologic and elemental (i.e. nutrient) system dynamics, with a particular focus on responses of nutrient fluxes to disturbance.

While much research has been done over the past 50 years on watershed ecosystems from a whole systems analysis standpoint, much less is known about internal mechanisms that control whole watershed systems responses. I will review a case study in a mountain watershed, where we investigated effects of removing near-stream understory vegetation and of natural blowdown of canopy trees on nutrient export to streams at the sub-watershed scale. The results from this study suggested that understory vegetation plays a relatively minor role in controlling nutrient export to headwater streams. Our results further suggested that nutrient uptake by canopy trees is a key control on nitrogen export in upland riparian zones, and disruption of the root-soil connection in canopy trees via uprooting can promote significant nutrient loss from the watershed ecosystem.

Bio(s):
J. Alan Yeakley is an Associate Professor of Environmental Science at Portland State University, in Portland, Oregon. Dr. Yeakley's research interests span ecosystem ecology and watershed hydrology, with a focus on riparian processes and urban ecology. He holds a BS in mathematics from Texas A&M-Commerce, an MS in environmental science from UT-Dallas and a Ph.D. in environmental science from the University of Virginia, where he was a presidential fellow. Alan has published articles in a variety of ecological science journals such as BioScience, Ecosystems, Ecology, Biogeochemistry and Landscape Ecology. He helped found the environmental science (ESR) undergraduate and masters programs at PSU, the Urban Ecosystem Research Consortium (UERC) of Portland/Vancouver, and is a member of the editorial board of ?coscience, an international journal of ecology. During this coming spring term, Alan will teach courses in Environmental Sustainability in Rosario, Argentina. For more information, please see http://web.pdx.edu/~yeakleya/alan.htm.

File attached or Link to Recording?
None



Date:
February 16, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Radu Popa

Title:
Xenobiology - The other side of life

Abstract:
"Xenobiology - Scientific discipline aiming to decipher life, its origin and evolution apart from its physical makeup"

Many dynamic systems including fire, fluid vortices, periodic reactions, chemical automata, crystals while growing, computer games, economy, society, GAIA and the Internet have features resembling life. Common sense tells us these are not alive. Yet, in the absence of a material-independent and quantifiable description of life we cannot tell how close these systems are to become alive. If we expect life to exist on other celestial bodies or to be simulated by computer modeling and algorithmic chemistry, we have to acknowledge that the concept of life is independent of the particular materials living entities are made of. In Xenobiology the essence of life is independent of things such as proteins, DNA, carbon or even water; this discipline considers that Earth's prebiotic chemistry was just one of the many possible frameworks on which life could have originated. The truly universal features of life relate to: energy
flow, self-control, departure from thermodynamic equilibrium, handedness, complexity, manipulating information and adaptive evolution. The ultimate goal of Xenobiology is to identify conditions allowing life to self-originate without design. This presentation reviews challenges and approaches when connecting energy flow with changes in complexity and the origin of genetic information. This approach requires connecting disequilibrium thermodynamics with changes in order and complexity, exploring strategies used by complex dynamic systems to evade deterministic chaos and identifying physical drivers of the evolution of dynamic systems. In the near future such knowledge will help model the link between energy dissipation and
changes in the organization of dynamic systems, and help explain general trends such as: the origin of life, macroevolution and the evolution of ecosystems toward climax.

Bio(s):
EDUCATION
California Institute of Technology and the Jet Propulsion Laboratory (Pasadena, CA) (Post Doc., 2002)
University of Cincinnatti, (OH) Environmental Microbiology (Ph.D., 2000)
The American University (Washington, DC) Evolutionary Biology (M.S., 1996)
University of Bucharest (Romania) Biology (B.A., 1983)

APPOINTMENTS
08/2005 => Present Tenure tracking, Associate Professor, Portland State University (Portland, OR).
2002-2005 Research Assistant Professor, University of Southern California (Los Angeles, CA)
1990-1994 Microbiologist, "E. Racovitza" Institute of Biospeleology (Bucharest, Romania).
1986-1990 Biologist, Central Institute of Biology (Bucharest, Romania)

SELECTED PUBLICATIONS
2006 Fisk M.R., R. Popa, O.U. Mason, M.C. Storrie-Lombardi, and E.P. Vincenzi, Iron-magnesium silicate bioweathering on Earth (and Mars?), Astrobiology, 6:48-68.
2006 Capone D., R. Popa, B. Flood and K.H. Nealson, Follow the nitrogen, Science, 312:708-709.
2005 Abboud R., R. Popa, V. Souza-Egipsy, C.S. Giometti, S. Tollaksen, J.J. Mosher, R.H. Findlay and K.H. Nealson, Low temperature growth of Shewanella oneidensis MR-1. Appl. Environ. Microbiol., 71:811-816.
2005 Nealson K.H. and R. Popa, Introduction and Overview: What do we know for sure? 1-24, Chapt. 1, In: Quantitative Approaches Towards Biogeochemistry: Processes, Scaling, and Interfaces, (L?ttge A. and R. Rye (eds.), American Journal of Science, Yale University, New Haven.
2005 Nealson K.H. and R. Popa, Metabolic diversity in the microbial world: relevance to exobiology, In: Gadd G.M., K.T. Temple and H.M. Lapin-Scott (eds.), 65th Symposium of the Society for General Microbiology, Micro-organisms and Earth Systems, Advances in Geomicrobiology, 151-171, Cambridge University Press.
2004 Popa R. and B.K. Kinkle, Controlled Mineralization of Pyrite by Thiomonas thermosulfatus str.51, Geomicrobiol. J., 21:193-206.
2004 Popa R., A. Badescu and B.K. Kinkle, Pyrite framboids as biomarkers for iron-sulfur systems, Geomicrobiol. J., 21:1-14.
2003 Fisk M.R., M.C. Storrie-Lombardi, S. Douglas, G.D. McDonald, R. Popa and A.I. Tsapin, Evidence of Biological Activity in Hawaiian Subsurface Basalts, Geochem. Geophys. Geosys, 4:1525-2027.
2002 Cox L., R. Popa, K.H. Nealson, and D. Bazylinsky, 2002, Organization of P-, S- and Fe-inclusions in a freshwater Magnetococcus, Geomicrobiol. J., 19:387-406.
2000 Popa R. and B.K. Kinkle, Discrimination among iron sulfide species formed in microbial cultures, J. Microbiol. Meth., 42:167 174.

File attached or Link to Recording?
None



Date:
February 9, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Lars Holmstrom

Title:
Receptive Field Models of Auditory Neurons

Abstract:
The modern quest to understand the physiology of sound processing and speech perception is often traced back to the work of Hermann von Helmholtz in the mid 19th century. While much has been learned since then, far more questions have been generated than have been answered. While Helmholtz was primarily focused on the mechanics of the ear, the modern auditory physiologist is often focused on the neural mechanisms at play. This is a daunting task considering the numerous auditory nuclei (functional collections of neurons) which are connected by a web of both ascending (toward the audio cortex) and descending (towards the ear) neural pathways. While it is generally understood that a systemic approach is required to fully understand this complex system of interactions, the state of the art is still primarily focused on making sense of how individual neurons respond to auditory stimulus. This talk will focus on some of the common techniques used for modeling the response characteristics of individual auditory neurons. It will focus primarily on Spectro-Temporal Receptive Field (STRF) models with special attention to the effects of experimental design on the fitting of the model. Examples will be provided of using these models in the context of researching the role played by neurons in the inferior colliculus (IC) of the mustached bat in the perception of social vocalizations.

Bio(s):
Lars Holmstrom is a PhD student in the Systems Science Department at Portland State University. Primary academic interests involve the use of statistical signal processing methodologies in the analysis of neural signals. His "other life" as a musician has led him in the direction of applying these tools to further our understanding of neural signal processing in the auditory system.

File attached or Link to Recording?
None



Date:
February 2, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Elizabeth Turgeon

Title:
Optimality vs. Resilience in an Agent-Based Evolutionary System

Abstract:
In evolutionary systems, there is a trade-off between optimality and resilience. Optimally-adapted populations show little variation in their descendants, thus making their descendants better able to compete in current conditions, while resilient populations preserve variation, trading this-moment competitiveness for the potential to withstand future, unknown threats. From observation of evolutionary systems, however, there appears to be a distinct advantage to optimality over resilience. Elizabeth Turgeon will be presenting an agent-based simulation that explores the trade-off between optimality and resilience in one evolutionary system.

Bio(s):
Elizabeth Turgeon is an engineering master's student, and has spent the past six years as a safety engineer, trying to prevent injuries, fires, and other unintended effects of manufacturing processes.

File attached or Link to Recording?
None



Date:
January 26, 2007

Location:
Harder House, Room 104

Time:
12-1 pm

Presenter(s):
Rich Jolly

Title:
Simulation of Information Sharing in Organizations

Abstract:
Game theory and agent based simulation have been used to study information sharing in organizations. In particular, the central tendency of organizational members to either freely share information, or keep it to themselves for their own personal gain, was studied in detail. First, the fundamental interactions are understood using a game theoretic perspective. Then, a simulation was conducted with the tool Netlogo where agents were assigned the fundamental property of either being a hoarder of information (unwilling to share) or a sharer of information. The effect of sharing on the organization was studied and it was found, as expected, that sharing greatly increases the overall information within the organization. The unexpected result is that agents who share tend to acquire more information than agents that hoard. This result is due to the synergy that develops between groups of agents who are sharing with each other building up the information levels greatly. It is also seen that the density of the agents is a critical parameter. As the density increases the probability increases that an agent is located near someone who has a large amount of information to share. The simulation results counsels organizations to use techniques to foster information sharing and discourage hoarding. This study has shown agent based simulation and a careful simulation methodology to be powerful tools in the study of organizational phenomena.

Bio(s):
Rich Jolly is a PhD candidate with the Systems Science and School of Business option. Rich's research interests are focused around using the systems perspective to improve the effectiveness of information technology in organizations and business. Rich also works at Intel Corporation in planning and market research for the server products group.

File attached or Link to Recording?
None



Date:
January 19, 2007

Location:
Harder House, Room 104

Time:
5-7 pm

Presenter(s):
Dr. Ftitjof Capra
Center for Ecoliteracy, Berkeley

Title:
The contemporary importance of systems ideas

Abstract:
He will talk briefly about the contemporary importance of systems ideas and we will have an informal, open-ended discussion with him.

Bio(s):
Dr. Fritjof Capra, Center for Ecoliteracy, Berkeley,is well known systems theorist, ecological activist, author, and filmaker.

File attached or Link to Recording?
None