Search Google Appliance


Seminar Archive 2009

 


 

FALL 2009 SCHEDULE

 


DATE: December 4, 2009

LOCATION: Harder House, Room 104

TIME: 12 noon - 1 pm

PRESENTER: William "Ike" Eisenhauer

TITLE: "Creating insanity in learning systems: addressing ambiguity effects of predicting non-linear continuous valued functions with reconstructabilty analysis from large categorically valued input data sets"

ABSTRACT: Being told to give two different, and potentially counter, responses to the same stimulus can set up a double bind in humans, leading to a type of insanity. So what how do you deal with it when it comes up quite frequently in modeling through simplification and removal of predictive variables?

In his current dissertation research Ike Eisenhauer is using reconstructability analysis to implement K-System, U-System, and B-System approaches to predict a continuously valued function through discrete categorically valued input variables [e.g. textual data]. One of the key issues is how to address the inability of K-Systems and U-Systems to allow the same input to give two different outputs, as well as how to report the performance of learning predictive systems which are trained to know that the multiplicity will exist.
This discussion session will consist of a quick overview of Ike's current work and then discuss the key point of: If a system has learned [been trained] that there are two or more different "correct" responses to a given stimulus, what should it report if it is only allowed to pick one response? Especially when it is "punished strongly" for not giving the other one, regardless.

BIO: William "Ike" Eisenhauer is an adjunct assistant professor in both the Systems Engineering and Engineering and Technology Management departments at Portland State University. His research interests include: Adaptive Belief Management, State Based Reconstructability Analysis, Shared Resource Constrained Data Envelope Analysis, Conflict Under Deceptive Irrationality, and Sustainable Quality Management Program Development. In addition to teaching, Ike is Chief of Systems Engineering for the U.S. Department of Veterans Affairs Portland VA Medical Center. His work there is focused on the systemic improvement of health care delivery. Prior to joining the DVA, Ike held positions at Wells Fargo in Risk and Loss Management and Equity Operations. He is an industry consultant in the areas of probability/uncertainty management, executive decision making, and benchmarking. His past clients have included US Bank, Hollywood Entertainment, and Multnomah County.

[return to top]


 

DATE: November 20, 2009

LOCATION: Harder House, Room 104

TIME: 12 noon - 1 pm

PRESENTER: Marek Perkowski

TITLE: "Towards robot theatre"

ABSTRACT: The talk will present the idea of futuristic robot theatre and work done towards it at the Intelligent Robotics Laboratory, Department of Electrical and Computer Engineering at PSU. After a short history of robot theatre from antiquity until 2008 we will present recent work on robot theatre in the world and at PSU, including two plays: ancient Korean folk tale "Hahoe Pylyshin" and "What's that? A Schroedinger Cat" or a debate between Einstein and Schroedinger Cat about quantum mechanics - an educational theatre. Several models of robot theatre will be discussed: animatronic theatre, interactive theatre and improvisational theatre. We will present the concept of generalized motions and universal event editor to edit robot motions, behaviors, lightings and automated events, as well as partial theories used to design such software tools, including regular expressions and spectral filtering theories. Next the concept of future interactive robot theatre will be presented, together with its underlying theories of pattern recognition and emotional robotics. Human-robot interaction based on recognition of human emotions and generating emotional robot behaviors as well as the method of constructive induction will be briefly discussed. Some ideas for future robot theatres will conclude the presentation. Our goals are to both create a model innovative robot theatre and a theory of robot theatre that would be similar to the theory of film or theory of interactive computer games. We believe that robot theatre will become a new art form and we are interested what are the basic questions related to the art of performing robots. We hope to have an interesting feedback to our ideas from the System Science oriented researchers.

BIO: Marek Perkowski obtained his M.S. degree in Electronics and Ph.D. Degree in automatic control from Institute of Automatic Control, Department of Electronics, Technical University of Warsaw, Warsaw, Poland. He studied also pure mathematics at University of Warsaw. In years 1981-1983 he was a Visiting Assistant Professor at University of Minnesota in Minneapolis and since 1983 he works for Department of Electrical and Computer Engineering at Portland State University where he is a full professor and director of Intelligent Robotics Laboratory. He worked for Cypress Semiconductor (co-author of WARP, the first FPGA compiler of VHDL), Intel Supercomputer, Sharp Microelectronics, GTE and other companies in areas of computer architecture, CAD tools for logic synthesis and image processing. Dr. Perkowski invented Kronecker Decision Diagrams and lattices and contributed to logic synthesis software that is used in US industry. In 1994 he worked for Machine Learning group in Wright Laboratories of U.S. Air Force applying logic decomposition as a machine learning approach to pattern recognition and continued this work on several grants. He is an author of more than 300 papers in CAD, logic synthesis, multiple-valued logic, machine learning, robotics and quantum computing. He had visiting professor and visiting scientist positions in the Netherlands, France, Japan and Korea. In years 2002-2004 he was professor in KAIST – Korean Advanced Institute of Science and Technology where he participated in research on humanoid robotics and quantum computing. He chaired the IEEE Technical Committee on Multiple-Valued Logic in years 2003-2005 and is currently chair of IEEE Computational Intelligence Society Task Force on Quantum Computing. His main current interests are in quantum circuits and algorithms, humanoid bipeds, emotional quantum robots, robotics for teenagers and Grover algorithm. He collaborates with many groups worldwide.

[return to top]


 

DATE: November 13, 2009

LOCATION: Harder House, Room 104

TIME: 12 noon - 1 pm

PRESENTER: Radu Popa

TITLE: "The complexity-independence of the origin of life"

ABSTRACT: It is often stated that the macroevolution of life is driven toward increased Complexity, and indeed, biosystems situated at higher evolutionary level show higher levels of Complexity. Yet, evidence also shows that some dynamic systems evolve toward lower entropy states, and not by increasing Complexity, but by increasing Organization. Organization is a parameter with two almost orthogonal components: Order and Complexity. Hence, it is possible for a dynamic system to experience changes in Organization in ways that do not elicit changes in Complexity. Whether Order or Complexity controls changes in Organization is dictated by the capacity of a system to store Meaningful information, and by the costs and payoffs of changes in Order or Complexity. This presentation analyzes transitions in the evolution of prebiotic systems (microevolution events) that are Complexity-independent. It is concluded that the actual driver of evolution is not the need for more Complexity, but the need to maximize the efficiency of energy dissipation.

BIO: Radu Popa has been an Associate Professor in the Biology Department at Portland State University since 2005. His research interests include microbial ecology and the origin and evolution of prebiotic systems (the origin of homochirality).

B.S. Biology - University of Bucharest, Romania
M.S. Evolutionary Biology - The American University, Washington DC
Ph.D. - Ecology - University of Bucharest, Romania
Ph.D. - Microbiology - University of Cincinnati, OH
Post Doc. - Caltech/JPL Pasadena, CA
Res. Prof. - University of Southern California, Los Angeles, CA

[return to top]


 

DATE: November 6, 2009

LOCATION: Harder House, Room 104

TIME: 12 noon - 1 pm

FACILITATOR: Joshua Hughes

TOPIC: "Generalists, specialists, and the best experts: Where do systems thinkers fit in?"

"GENERALIST / SPECIALIST: A generalist is someone who has studied a little bit of everything, and in the end knows nothing well in particular. By contrast, a specialist is someone who has studied a single subject, and as a consequence does not even know his own subject, because every item of knowledge is related to other components of the whole system. The good scholar or scientist--like the good chef, manager, clinician, or orchestra conductor--is an expert in one field or craft, and knowledgeable in many. Like a mouse, he can explore the details of a terrain; and, like an owl, he can also soar to get a good view of the landscape--mice and all. He is capable of learning new subjects as needed, as well as placing every particular subject in a wide context and a long-term perspective. He is thus open to multiple inputs and capable of multiple outputs. In sum, the best expert is the specialist turned generalist. This holds in all fields of thought and action, particularly in philosophy." -- Mario Bunge, Philosophical Dictionary

Bunge's definitions of the generalist, the specialist, and the "best expert" are thought-provoking (and may provoke other responses as well). Are systems practitioners, analysts, and theorists generalists, specialists, or the best experts? Because systems science concerns itself with general theories (e.g. graph theory, information theory, control theory, game theory, etc.) that can be applied to a wide range of problems, it appears to be a generalist field; but systems science has its own contributors, jargon, and history and is not widely studied (at least in the U.S.), and so appears to be a specialist field as well. And yet many of the early contributors to the systems project such as von Bertalanffy, Boulding, Wiener, and Ashby did indeed fit Bunge's definition of the best expert, as all were specialists turned generalists. Since systems science is mostly taught at the graduate level, perhaps Bunge's position is an implicit assumption in the systems field.

You may not agree with all of Bunge's assertions (or the conjecture above), but it is clear the views of the mice and the owls are needed for most (if not all) problems. Is the systems view that of the owls or that of mice in owl clothing? The answer may be fuzzy and a good starting point for our discussion. Perhaps a more interesting question is this: How we can use systems thinking to improve our problem solving abilities? A quick look at the jobs graduates of the PSU Systems Science Graduate Program have gone on to (http://www.pdx.edu/sysc/resources-jobs) makes it clear that systems principles are applicable in all kinds of fields. It is also clear that systems science can be useful for framing and solving global problems related to economics, energy, climate, and politics. So whether generalist or specialist--or whether one can meet the criteria Bunge requires of a "best" expert--what roles can a systems thinker fill?

Here are a few questions to get the discussion going:

  1. Are you interested in being a general problem solver, or do you have a specific (i.e. specialized) problem you'd like to solve using systems thinking?
  2. Can you describe an instance when your knowledge of systems science gave you an insight you would not otherwise have had?
  3. What roles can systems theorists, analysts, and practitioners play in national and global debates?
  4. Do (or will) the public, politicians, and other experts accept systems thinkers as experts?
  5. Can (or do) systems practitioners and theorists act as liasons between specialists or between specialists and the public?
  6. Can you think of a field or a problem that is not being considered from a systems perspective but should be?
  7. (Extra credit) Can you think any field in which systems science would not be useful?

This discussion can also be an opportunity for new students to ask questions about the systems field and discuss what they hope to gain with systems science knowledge, and for other students, graduates, and faculty to share their insights and experiences about the systems field and what they have gained from their systems science knowledge.

BIO: Joshua Hughes is a second year, core-option PhD student and graduate assistant in the PSU Systems Science Graduate Program. He is beginning research with George Lendaris on contextual learning and experience-based identification and control; he is also collaborating with Martin Zwick on a few papers that show how systems theories might provide insights into some contemporary problems. He is interested in information theory, cybernetics, reconstructability analysis, neural networks, fuzzy logic, catastrophe theory, game theory, and many other things.

[return to top]


 

DATE: October 30, 2009

LOCATION: Harder House, Room 104

TIME: 12 noon - 1 pm

PRESENTER: Heejun Chang

TITLE: "Water as a complex system: understanding the dynamics in a changing environment"

ABSTRACT: The water resources system is constantly evolving over space and time at a range of scales. Human-induced climate change and land development are probably two major driving forces of water resource system changes. However, the impacts of such changes are region specific, which depend on watershed characteristics such as topography and geology. Numerical simulation models are useful tools for understanding the system dynamics by allowing the multiple interactions of system components. I will introduce case studies of the Pacific Northwest that examine how changing climate and population growth affect regional water resources at multiple spatial and temporal scales and explain the major determinants of such changes in the system. A combination of a GIS-based hydrologic model and a hydroeconomic model is used for integrated environmental change impact assessment.

BIO: Heejun Chang is an Associate Professor of Geography at Portland State University, in Portland, Oregon, USA where he teaches courses in physical geography, hydrology, climate and water resources, global water issues and sustainability, GIS for water resources, and spatial quantitative analysis. His research areas include impacts of climate variability and change on regional water resources, land cover change and water quality, use of geospatial technology for hydrology and water resources, and urban flooding in Monsoon Asia. Professor Chang’s work has been funded by the Sustainable Water Resources Program at the Ministry of Education, Science and Technology of Korea (Technology for Climate Impact Assessment), the US National Science Foundation (Urban Water Quality), NOAA (climate change and urban water demand), NRCS (spatial database development), and the James F. and Marion L. Miller Foundation (Sustainable Water Resources Management, Hydrologic Ecosystem Services Dynamics, Coupled Carbon and Water Cycle in Urban Areas). His recent publications appear in such interdisciplinary, international journals as Climatic Change, Climate Research, Hydrological Processes, International Journal of Climatology, Journal of Environmental Management, River Research and Applications, Science of the Total Environment, and Water Research. Chang is currently a representative for the Willamette River Basin in UNESCO’s HELP (Hydrology, Environment, Land and Policy) program. He holds a Ph.D. in Geography from the Pennsylvania State University. For more information about Chang’s research, please visit http://www.web.pdx.edu/~changh/research.html.

[return to top]


 

DATE: October 23, 2009

LOCATION: Harder House, Room 104

TIME: 12 noon - 1 pm

PRESENTER: Richard Beyler

TITLE: "Doing the history of science and the suspension of belief"

ABSTRACT: Samuel Taylor Coleridge (1817) described the "suspension of disbelief" as a sort of bargain between the author and the audience necessary to creative literature. Conversely, one might describe the "suspension of belief" as a necessary element of doing the history of science. In modern civilization, science counts as the social institution which delineates the boundaries of knowledge per se, as opposed to belief, opinion, etc. We might describe this as the cultural myth of science--not in the sense of judging its truth or falsehood, but rather in the sense of its being foundational and largely unquestioned in modern society. Yet a historical account of how science developed requires that we suspend our current state of knowledge as a relevant factor, unless we are prepared to claim, teleologically, that this current state of knowledge played a causative role in past developments. Two aspects of the cultural myth of science have seemed to me particularly persistent and needing of suspension in order to create a cogent historical account: 1) an almost overwhelming tendency to interpret historical significance (exclusively) from the perspective of the known "winners"; 2) Uncritical belief in pure science or "knowledge for its own sake" which casts any kind of economic, political, or other "external" involvement as necessarily deterimental to the scientific enterprise.

BIO: Richard Beyler received his Ph.D. in history of science from Harvard University in 1994. After post-doctoral fellowships in Berlin and in Washington, he came to Portland State University in 1996, where he is associate professor of history. His teaching fields are history of science, European intellectual history, and German history. His current research follows two main tracks: biophysics in the 1920s and 1930s, and the political realignment of German scientific institutions before and after World War II.

[return to top]


 

DATE: October 16, 2009

LOCATION: Harder House, Room 104

TIME: 12 noon - 1 pm

PRESENTER: Tad Shannon

TITLE: "Linguistic Adaptive Critics and Human in the Loop Dynamic Programming"

ABSTRACT: Adaptive critic methods for approximate dynamic programming are a subset of reinforcement learning techniques developed over the past several decades. Actor-critic methods segment control system design problems so that assemblages of computational devices can efficiently find near optimal control policies for complex systems. While much effort has been devoted to designing autonomous adaptive critics, relatively little work has been done on human interaction with actor-critic systems. This talk will provide a taxonomy for adaptive critic systems, review developments in the use of linguistic reinforcement signals in such systems, and suggest a variety of options for human interaction with actor-critics.

BIO: Tad Shannon is an Assistant Professor of Theatre-Dance at Western Oregon University in Monmouth, Oregon. He has been the resident designer with Portland’s Do Jump Movement Theatre since 1992, designing the lighting and rigging for national tours in 2000, 2001, and 2007. He received a Theatre LA Ovation award for best lighting design for Do Jump’s Openings and Doors in 2002. His design work has been seen with many regional dance and theatre companies in Portland and San Francisco. He received his Ph.D. in Systems Science from Portland State University in 2007.

[return to top]


 

DATE: October 9, 2009

LOCATION: Harder House, Room 104

TIME: 12 noon - 1 pm

PRESENTER: Mark Bedau

TITLE: "Beyond Biobricks: synthesizing synergistic biochemical systems from the bottom-up"

ABSTRACT: Engineers who attempt to discover and optimize the behavior of complex biochemical systems face a dauntingly difficult task. This is especially true if the systems are governed by multiple qualitative and quantitative variables that have non-linear response functions and that interact synergistically. The synthetic biology community has responded to this difficulty by promoting the use of "standard biological parts" called "BioBricks", which are supposed to make biology into traditional engineering and enable engineers to "program living organisms in the same way a computer scientists can program a computer". But the BioBricks research program faces daunting hurdles, because the nonlinearity and synergy found throughout biochemical systems generates lots of unpredictable emergent properties. This talk describes an alternative vision of how to engineer complex biochemical systems, according to which we would refashion engineering to fit biology (rather than the other way around). The resulting method (termed "Predictive Design Technology" of PDT) is a robot- and computer-driven automatic and autonomous implementation of traditional Edisonian science. The PDT method is described and illustrated in application to a number of practical biochemical design tasks, including (2) optimizing combination drug therapies, (2) optimizing cargo capacity of liposomes that self-assemble from complex amphiphile mixtures, (3) optimizing the liposomal formulation of insoluble drugs, and (4) optimizing in vitro protein expression.

BIO: Mark A. Bedau is Professor of Philosophy and Humanities at Reed College in Portland, Oregon, Co-Founder of the European Center for Living Technology (ECLT), Partner in the EU-funded Programmable Artificial Cell Evolution (PACE) program, Co-organizer of the Eleventh International Conference on the Simulation and Synthesis of Living Systems (Artificial Life XI), and Visiting Professor, Ph.D. Program in Life Sciences: Foundations and Ethics, European School of Molecular Medicine. He is the coeditor of Emergence: Contemporary Readings in Science and Philosophy and Protocells: Bridging Nonliving and Living Matter, both published by the MIT Press in 2008.

[return to top]


 

SPRING 2009


Date: June 5, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Rich Jolly

Title: The Role of Feedback in the Assimilation of Information in Markets: Applications to Prediction Markets

Abstract: “If only HP knew what HP knows, we would be three times more productive” - Lew Platt, while CEO of Hewlett-Packard

Leveraging the combined knowledge of an organization is an ongoing challenge and has given rise to the field of knowledge management. Yet, despite spending enormous sums of organizational resources (time and money) on IT (Information Technology) systems, executives recognize there is much more knowledge to harness – as expressed by Lew Platt’s comment above. Prediction markets are emerging as one tool to help extract, and make operational, that extra knowledge. Yet, prediction markets, like other markets, are susceptible to pathologies which compromise their accuracy (e.g. bubbles and crashes). This makes their use problematic for organizations.

The goal of this research is to study aspects of the feedback relationships in markets. A simplified form of a commonly used organizational prediction market will be studied. Some preliminary results of system simulations will be discussed in the seminar.

Bio: Rich Jolly is a PhD candidate in Systems Science and Business Administration at PSU. His interest is in the application of systems science tools and methodologies to business problems--in particular the flow of information in organizations. Rich currently works full time at Intel doing strategic marketing for server products.

File attached or Link to Recording?

Coming Soon

[return to top]


 

Date: May 29, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Chad Wiener

Title: R.G. Collingwood on Object and Method in History

Abstract: Historical knowledge seems problematic in several ways. First, the knowledge would be of particular events at a specific time and place. The standard scientific method of finding classes and constructing hypotheses to predict phenomena will not work for historical knowledge. Second, although historical knowledge is empirical, it is not clear that it can be confirmed or falsified by experiment. I will show how Collingwood argues historical knowledge is possible and defend his thesis that all history is the history of thought. His argument is based on the claim that the object of history is human actions, and human actions contain as one of its constituent elements thought. I will explain what a human action is for Collingwood and how we can have knowledge of such actions in the past.

Bio: Chad Wiener received his Masters in Social Science from the University of Chicago and his PhD in Philosophy from the University of Georgia. He is currently working on scientific inquiry in Plato and Aristotle. He specifically works on Aristotle's biology and its connection to Aristotle's metaphysics and logic. He also dabbles in and intends to publish on methodology in history, especially on the thought of R.G. Collingwood.

File attached or Link to Recording?
Coming Soon

[return to top]


 

Date: May 22, 2009

Location: Harder House, Room 104

Time: 12-12:50

Presenter: Bobby Cochran

Title: WHO´S MAKING DECISIONS IN MARKETS FOR ECOSYSTEM SERVICES?: Using Network Science to Reveal New Patterns

Abstract: Market-based approaches to meeting environmental goals present an exciting governance mechanism with a potential to bridge these divides between urban and rural communities, technocratic and deliberative forms of governance, and the equity issues of who is best able to produce environmental benefits. Using social network analysis, this research provides empirical support that the power of environmental markets as a bridge stems not from their conception as places of economic exchange, but as forums for new social exchanges as diverse stakeholders negotiate new institutional designs.

Bio: Bobby Cochran is an Adjunct Professor in the Dept of Urban Studies and Planning at Portland State and the Environmental Marketplace Analyst at Clean Water Services in Washington County. He has his PhD in Urban Studies and Masters in Conflict Resolution and Public Policy.

File attached or Link to Recording?
Coming Soon

[return to top]


 

Date: May 15, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Joshua Hughes

Title: Is the panarchy adaptive cycle a special case of the cusp catastrophe?

Abstract: Comparison of the panarchy adaptive cycle, a general model for human and natural systems, with the cusp catastrophe of catastrophe theory suggests that the adaptive cycle can be considered a special case of the cusp catastrophe. Both the adaptive cycle and the cusp catastrophe have been used to model various ecological, economic, and social systems in which slow, small, continuous changes in one or two control variables produce a fast, large, discontinuous change in system behavior. Use of the panarchy adaptive cycle, the more recent of the two, has so far been limited to that of metaphor, but the adaptive cycle still provides rich explanatory power and philosophical insight for many living systems. The cusp catastrophe, while often used as a metaphor, has been derived from topology and so is capable of being used much more rigorously. By using the constrained control variables from the adaptive cycle as parameters in the behavior equation for the cusp catastrophe, a cycle very similar to the adaptive cycle is constructed. Where the constructed cycle differs from the adaptive cycle is where the adaptive cycle is least well-defined, and several ways for eliminating this discrepancy are discussed. Considering the panarchy adaptive cycle to be a special case of the cusp catastrophe may provide direction for more rigorous and more general applications of the adaptive cycle, thereby extending its usefulness in guiding sustainability efforts, the primary purpose for which it was created.

Bio: Joshua Hughes is a first year, core-option PhD student in the PSU Systems Science Graduate Program. He is interested in information theory, cybernetics, reconstructability analysis, neural networks, fuzzy logic, catastrophe theory, game theory, and many other things.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p24958174/

[return to top]


 

Date: May 8, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenters: Rajesh Venkatachalapathy and Robin Fenske

Title: A Discussion about Science and Advocacy

Abstract: As scientists we strive for objectivity and developing unbiased knowledge in our research, but as human beings we often feel compelled to try to influence the systems we study...which can introduce bias. How do we/should we strike the proper balance between scientific objectivity and advocacy--trying to make the world a better place? Scientific ideas can be shared, but at what point do they become biased and inaccurate? Come and discuss 1) your own challenges with sharing your ideas while minimizing unnecessary bias, and 2) thoughts on how much advocacy of scientific ideas is too much or not enough (i.e. should scientists be advocates? At what point do we dilute, mystify, or popularize science too much? Perhaps there should be an interplay between objectivity (science) and functionality (advocacy), but the two ideas should remain separate).

Bio(s): Rajesh is currently a Systems Science student and is working with Martin Zwick on comparing RA with other Machine Learning Models. Robin is a second year PhD student in the Systems Science Graduate Program. She is a Core Student. She is interested in applying Systems Science to Sustainability in Food Systems, and Human Decision Making.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p13716382/

[return to top]


 

Date: May 1, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Jess Laventall

Title: It's the System: Crack Dealing and Systems Dynamics

Abstract: Crack dealing gangs may not be the first thing that comes to mind when thinking about systems dynamics. Yet, that is exactly what is described in Levitt and Dubner’s best-selling book, Freakonomics. These two economists devote a chapter to explaining why drug dealers still live with their mothers. Based on the study they reference, original research conducted by their colleague Sudhir Venkatesh, who actually infiltrated a notorious drug gang to collect data over a four year period, we find some revealing facts that inform and make apparent the behaviors of a dynamic system a drug gang operated under given the conditions described.
This presentation provides an overview of a systems dynamics approach to how a crack dealing drug gang operated in an inner-city environment. A model was constructed to explore the interaction effects of gang violence, finances and member recruitment. The volatility of the situation these gangs faced in their environment offers and excellent observable system.

Bio: Jess Laventall is a graduate certificate student in the Systems Science Department at Portland State University. He is co-founder of American Choice Modeling, an advanced market modeling and simulation firm. His background includes a wide variety of quantitative approaches to marketing and marketing research.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p80733809/

[return to top]


 

Date: April 24, 2009

Location: Harder House, Room 104

Time: 12-12:50

Presenter: Lonneke Eeuwes

Title: Bioelectric fields in aquatic organisms: a hidden information source?

Abstract: All aquatic organisms are surrounded by a weak bioelectric field that consists of a direct current (DC) and an alternating current component (AC). The AC component is caused by (ventilatory) movements, whereas the DC component results from biochemical processes such as osmoregulation. A variety of factors modulate osmoregulation, which in turn causes the bioelectric fields to be dynamic in both strength and shape. Since bioelectric fields of electroreceptive fish exceed their behavioral detection tresholds, it is implied that information herein could be used by conspecifics.

Bio: My research interest is focused on how sensory information is processed by the neural system, especially in a complex (sensory) environment. I received a M.Sc. in Neurobiology/Zoology from Utrecht University (The Netherlands), where I did my PhD research as well. At the moment, I work as a postdoctoral research fellow at WSU on encoding of complex sounds in the mammalian auditory system.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p71075510/

[return to top]


 


Date: April 17, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Olgay Cangur

Title: Modeling Subprime Mortgage Delinquency, Termination and Loss

Abstract: The mortgage industry is facing a very challenging environment. Declining house prices have surfaced the importance of delinquency, loan default and loss predictions. Simple models of prepayment behavior are no longer applicable. Investors, originators, servicers and regulators are in need of accurate predictions for their portfolio of interest. My research focused on two topics relevant to modeling residential mortgages.

The first topic provided a framework for modeling delinquencies, prepayments, defaults and losses that represents an enhancement over previous studies. A total of nine loan payment statuses were used (current, thirty days late, sixty days late, ninety days late, early foreclosure, late foreclosure, real estate owned, paid in full, and terminated with loss). This framework was compared to the previous framework discussed in the literature that used seven statuses.

The second topic applied reconstructability analysis (RA) to 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 study explored the performance of RA versus logistic regression (LR). It also explored the hypothesis that inclusion of RA suggested interaction effects improves the accuracy of the LR. All three methods were compared.

The first topic's result made two unique and important contributions to the mortgage management literature. First, it determined that nine-state framework yields more accurate results compared to the seven-state framework. It also introduced a new state 'terminated with loss' that enabled the framework to predict losses. The second topic's results confirmed that RA was helpful in detecting and suggesting interactions effects that can be utilized in the logistic regression models within the payment model framework. However, the research could not conclude that RA yields better results than LR as a standalone prediction methodology, due to computational constraints and methodological limitations.

Bio: Olgay is currently working as a research analyst for Wilshire Credit Corporation in Beaverton, Oregon. He is developing a mortgage payment model for predicting future mortgage loan behavior that is similar in some ways to the models described above. The Wilshire model is used for risk assessment and pricing of mortgage loans. He is also pursuing his PhD in the Systems Science Graduate Program at Portland State University. His dissertation topic is described above. His research interests include data mining, optimization, and dynamic modeling.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p50687609/

[return to top]


 


Date: April 10, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter(s): Ralf Juengling

Title: Reconstructing Piecewise Polynomial Functions with Leclerc's Algorithm

Abstract: In signal reconstruction one attempts to recover a "true signal" from noisy data. Ingredients for a successful approach include models of the true signal, sensors, and noise processes at differentstages of data acquisition. Reconstruction of smoothsignals is typically done by convolution. Reconstructionof signals with discontinuities can, in general, not bedone so cheaply.

Algorithms for reconstruction of different classes ofpiecewise smooth signals have been proposed. In 1989 Leclerc, motivated by work in image analysis, published an algorithm for reconstructing two-dimensional,piecewise polynomial signals from noisy data. Leclerc's algorithm features remarkable adaptivity: in addition torecovering regions over which the true signal is smooth, it reconstructs a signal with variable polynomial orderand variable noise parameters, each chosen on a per-regionbasis. While the algorithm is very powerful, it is alsocomputationally very expensive and, it seems, has not been adopted by other researchers since its invention.

In the first part of my talk I derive Leclerc's algorithm, which is cast in the form of an unconstrained optimization problem with a discontinuous objective function. A continuation method is devised to enable numerical optimization algorithms for this problem. In the second part I discuss my efforts to find an optimization method that would turn Leclerc's algorithm into practical tool for reconstruction.

Bio: Ralf Juengling is Portland State graduate student in the Computer Science department since 2004. Ralf's research interests are in machine learning and computer vision.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p50467851/

[return to top]


 


Date: April 3, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Bill Bloom

Title: Theory grounded in data

Abstract: Much research in Systems Science is quantitative in nature, relying upon applications of advanced mathematics, including statistics, calculus and similar disciplines. But the discipline is also amenable to qualitative research methodologies. One such methodology is grounded theory, which emphasizes the generation of "theory grounded in the data" using qualitative rather than quantitative techniques. Bill Bloom will lead a discussion on the application of grounded theory methodology within the business arena, using as an example his dissertation study of why some domestic manufacturers adapt successfully to the threats of global competition while others do not.

Bio: Bill Bloom has been in the SySc Ph.D. program since 2004, pursuing the business option. He practiced law in the Portland area for 20+ years before obtaining his MBA from PSU in 2001, and has been teaching business courses in Portland and overseas for the past several years. He is interested in the systemic aspects of globalization, in particular the dynamics of offshoring manufacturing jobs overseas in pursuit of low wages.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p58536053/

[return to top]


 

 

WINTER 2009



Date: March 13, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Robin Fenske

Title: What is your favorite tangible/physical metaphor related to Systems Science?

Abstract: Hoil and I were talking about metaphors related to different fields of study, and how these metaphors can be used to popularize or synthesize or iconify ideas in these fields of study. For example, the ideas/images of Bull and Bear markets are iconic to economics.

There are different levels of metaphors, more abstract and more concrete. For example, the image of the Supply and Demand curves are sometimes used as a metaphor or icon in economics, but it is more abstract (when not implemented) than the Bull and Bear ideas So maybe Systems Science also has some helpful and fun metaphors.

Try to limit yourself to a more concrete metaphor, and not a meta-metaphor, meaning, try to think of a metaphor that can be represented by a real world thing (like a butterfly effect) not a more abstract metaphor (e.g. not Zwick's Function and Structure).

More examples:
We'd like to talk about things like:
Butterfly effect
Maxwell's Demon
Slime mold
Cornucopia

We wouldn't like to talk about things like:
Emergence
Wholism and Holism
A Level and B Level
Hierarchy


Please bring your ideas to seminar and we'll compare and discuss.

Bio(s): Robin Fenske is a full time second year PhD student in the Systems Science Graduate Program at PSU. She is a Core Student. She is interested in applying Systems Science to Sustainability in Food Systems, and Human Decision Making. She is planning on taking Comprehensives in Spring 2010.
She is a research assistant for the Winter 2008 through Winter 2009 quarters with Dr. Wayne Wakeland and Dr. Mellie Pullman on a "Food Delivery Carbon Foodprint" inter-departmental research grant. This research is on sustainable food purchasing within institutions, exploring institutional buyers’ demographics, values, motivations, and decision-making in the supply chain. She is also developing her own research. She is interested in understanding consumer behavior, intention, and assumptions, and how this understanding can help promote sustainable food systems and lessen rampant consumerism. She is also very interested in applying Qualitative research methods to her research topic. She holds a Bachelor of Science from The Evergreen State College in Washington, and has professional experience as an energy economics research analyst.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p55669871/

[return to top]


 


Date: March 6, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Ed Ramsden

Title: Mechanics and Implementation of a Systems Dynamics Simulation Engine

Abstract: System dynamics models can be developed using either general-purpose programming languages such as 'C++', or specialized modeling tools such as Vensim, Stella, or Powersim. The discussion this week will be on the
development of a simple system dynamics modeling tool, and will encompass the topics of data structures and algorithms used 'under the hood' in this type of software.

Bio: None

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p20739208/

[return to top]


 


Date: February 27, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Rajesh Venkatachalapathy

Title: In silico Experiments and Simulation

Abstract: In silico experiments have become common place practice in all areas of science as an accepted way of discovering knowledge. However there is no rigorous foundation on which the methods and practice stand. Our discussion this week will be on gathering opinion on what is the right and the wrong way to do simulation and possibly seek patterns of implicit definitions that people use in their respective fields.

Bio: None

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p47116900/

[return to top]


 


Date: February 20, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: James Kar

Title: Reconstructability analysis (RA) as an alternative modeling technique for time-series prediction

Abstract: RA is based in information-theory and uses the principle of maximum (Shannon) entropy to fit models. Specifically, we explore RA models for predicting stock market returns (S&P500). Using thirteen (13) macroeconomic and financial variables, we find that according to the BIC model selection criterion the best model uses only one predictor. For the sample sub-period from January 3, 2003 to October 15, 2008, the best model uses DAX as the predictor. However, for the sample sub-period from January 3, 2002 to October 15, 2008, and for the entire sample period from January 3, 2001 to October 15, 2008, the best model uses FTSE instead. This signals possible changes in the underlying (data) structure of the stock market over time. The surprising result is that, according to BIC, no model is predictive beyond 1-day ahead forecasting.

Bio: James is a practicing certified financial planner (CFP) focusing on global investment and stocks and options trading. He had taught a variety of finance courses at Portland State University, such as Investment, and Equity Valuation. He has an undergraduate degree in business (finance) and a master's degree in taxation. Currently, James is working on his PhD in Systems Science/Finance where he specializes in financial market forecasting.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p23363363/

[return to top]


 


Date: February 13, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Martin Zwick

Title: Reconstructability Analysis in Biomedical Research

Abstract: This talk will briefly explain reconstructability analysis (RA), a graphical modeling methodology developed in the systems community from the early work of Ashby. RA is based in information theory and graph theory and both overlaps and augments more widely known machine learning and statistical methods such as log-linear models and Bayesian networks. The talk will report on the use of RA in a recent bioinformatics study of human gene (SNP)- disease (diabetes) association and epistasis, and will mention some other biomedical applications.

Bio: Martin Zwick is a core faculty member in Systems Science.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p11503839/

[return to top]


 


Date: February 6, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Rod Walker

Title: Management Training Simulations Using System Dynamics Models

Abstract: In the first edition of The Fifth Discipline, written over 18 years ago, Peter Senge made a strong case for "microworlds" business simulations which allow managers to "learn by doing". Despite his endorsement and their intuitive appeal, successful simulations of this type are still relatively rare. In the last 5 years, we have successfully implemented 4 large online training simulations, all built around iThink/Stella models of relevant parts of the client?s business. Two of these simulations were created for executive training at Fortune 500 firms. We have seen strong acceptance for these simulations, and the internal system dynamics models provide a way to generate the rich, realistic experiences that are critical for this type of learning. This presentation will discuss some of these simulations, their general structure, key implementation considerations, and important characteristics.

Bio: Rod Walker is a current student in Systems Science, and is a management consultant with a concentration in business dynamics. He holds BSEE and MBA degrees from the University of Texas at Austin. After 25 years as an engineer, manager, and executive at Texas Instruments and Compaq Computer, Mr. Walker left Compaq in 1998 to begin consulting. As a VP at Compaq, he participated in several corporate-wide projects utilizing systems modeling experts with McKinsey & Company. The breakthrough nature of those projects prompted the inclusion of business simulation modeling as part of his emerging consulting practice.

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p40325441/

[return to top]


 


Date: January 30, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter(s): Christof Teuscher, Portland State University, Department of Electrical and Computer Engineering
http://www.teuscher.ch/christof

Title: Computers: Quo Vadis?

Abstract: Since the beginning of modern computer science some sixty years ago, we are building computers in more or less the same way. Silicon electronics serves as a physical substrate, the von Neumann architecture provides a computer design model, while the abstract Turing machine concept supports the theoretical foundations. However, the landscape of computing machines and computing paradigms is changing. The reasons are diverse and I will start this talk by highlighting the major trends and challenges. I will outline my visionary and long-term research efforts to address the grand challenge of building, organizing, and programming future computing machines. I will delineate potential solutions on how these challenges might be addressed. Self-assembled nano-scale electronics, cellular automata (CAs), and random boolean networks (RBNs) will serve as a simple showcase. Last, I will discuss the need for novel paradigms and unconventional solutions to address complexity issues, to obtain self-adaptation, self-(re)configuration, self-repair, and to build more adaptive, cognitive, and robust machines.

Bio: Christof Teuscher holds an assistant professor position in the Department of Electrical and Computer Engineering (ECE) at Portland State University and an Adjunct Assistant Professor appointment in Computer Science at the University of New Mexico (UNM). He obtained his M.Sc. and Ph.D. degree in computer science from the Swiss Federal Institute of Technology in Lausanne (EPFL) in 2000 and 2004 respectively. In 2004 he became a postdoctoral researcher at the University of California, San Diego (UCSD), in 2005 a distinguished Director's Postdoctoral Fellow at Los Alamos National Laboratory, and in 2007 a Technical Staff Member. His main research interests include emerging computing architectures and paradigms, biologically-inspired computing, complex & adaptive systems, and cognitive science. Teuscher has received several prestigious awards and fellowships. For more information visit: http://www.teuscher.ch/christof

File attached or Link to Recording?
http://psuniv.na3.acrobat.com/p40269281/

[return to top]


 


Date: January 23, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: Dr. Christopher Joel Dubay

Title: Systems Biology for Health & Disease

Abstract: None

Bio: Dr. Dubay's primary research interest is in complex genetic diseases (e.g. diabetes, hypertension, etc.) and the dissection of their genetic basis in model systems and eventually humans.

I have concentrated on the use of bioinformatic tools to aid in the scaling up of experiments and analyses required to address the isolation of multiple genetic loci present in complex traits. I currently direct the Oregon National Primate Research Center Colony Demographics & Informatics unit, which provides bioinformatic tools and genetics resources to support colony management and our research on a wide variety disease models for translational research.

I hold joint appointments in the departments of Medical Molecular Genetics and Medical Informatics and Clinical Epidemiology in the Oregon Health & Science University School of Medicine, where I support research and training in bioinformatics and related subjects. I am interested in developing and evaluating systems for clinical and research genetic analysis laboratories, such as laboratory information systems, and translating our new biological knowledge into clinical interventions. I am a founder of Genetic Information Management Systems, and an information management consultant for Proteogenix Inc, as well as a board member of the personal genomics firm Iverson Genetics.

File attached or Link to Recording?
None

[return to top]


 


Date: January 16, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenters: Wayne Wakeland and Hoil Kang

Title: Web-based Remote Access Curriculum Delivery

Abstract: A discussion and demonstration of synchronous web-based remote access course delivery technology, including thoughts about why this technology will become increasingly important in the future.

Bio: 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?
http://psuniv.na3.acrobat.com/p40983065/

[return to top]


 


Date: January 9, 2009

Location: Harder House, Room 104

Time: 12-1 pm

Presenter: None

Title: Discussion on Systems Science for Undergraduates at PSU

Abstract: None

Bio: None

File attached or Link to Recording?
None

[return to top]