Systems Science Seminar Series
SUMMARY: The Systems Science Seminar Series covers a wide-range of topics, providing an opportunity for presenters to share and attendees to become exposed to latest and often cutting-edge research from different fields and disciplines. Agent-based simulation, artificial intelligence, artificial life, genetic algorithms, machine learning, neural networks, signal processing, social networks, system dynamics, and science itself are just a few of the many diverse topics that have been presented, all in an informal environment where questions and discussion are encouraged. The Seminar Syllabus is also available here.
QUESTIONS AND COMMENTS: If you have questions or comments, would like to present a seminar, or if you have suggestions for topics or speakers, please email Garry Sotnik at gsotnik@pdx.edu.
Past seminar links and records can be found on our Seminar Archive page
SPRING 2012 SCHEDULE
- 04/06/12 - B. Max Grad, Systems Science PhD Student, PSU, Problem Solving and Information Processing in Human Systems. [Announcement] [Elluminate Recording]
- 04/13/12 - Lee Hullender Rubin, Post Doctoral Research Fellowship, Oregon College of Oriental Medicine, "Traditional Chinese Medicine Challenges in Research". [Announcement] [Elluminate Recording]
- 04/20/12 - Sergio Antoy, Professor of Computer Science, PSU, "Programming with Narrowing". [Announcement] [Elluminate Recording]
- 04/27/12 - Stephen Sharvais, Associate Professor of Management Information Systems, Eastern Washington University, "Satisficing vs Exploring in Multistage Processes" [Announcement] [Elluminate Recording]
- 05/04/12 - Bart Massey, Associate Professor of Computer Science, PSU, "Understand "Inform 7" as an Ontological Modeling Language". [Announcement]
- 05/11/12 - Shelby Anderson, Assistant Professor of Anthropology, PSU, "Ceramic Sourcing and Social Networks in Northwest Alaska". [Announcement] [Elluminate Recording]
- 05/18/12 - Mehmet Vurkac, Assistant Professor, Department of Electrical Engineering & Renewable Energy, Oregon Institute of Technology "Bayesian and Related Methods: Techniques Based on Bayes' Theorum [Announcement]
MVBayesPresentation.ppt [Elluminate Recording] 
ExtendedMarslandExampleForNaive-Bayes.xlsx - 05/25/12 - Jeff Fletcher and Martin Zwick, Professors of Systems Science, PSU, "Levels of Altruism"
- 06/01/12 - Dick Hamlet, Professor Emeritus of Computer Science, PSU, " The Intersection between Science and Computer Science is Almost Empty".
- 06/08/12 - Feng Liu, Assistamt Professor of Computer Science, PSU, "".
- 06/10/12 - Classes end
DATE: Friday, April 6th, 2012, 12:00 - 12:50 PM
PRESENTER: B. Max Grad (a.k.a. Deadletter)
TITLE: Problem Solving and Information Processing in Human Systems
SUMMARY: A system is a unit with attributes in relation to its environment, and sub-units whose interactions manifest that attribute. Problem Solving, Bloom's Taxonomy, Individual and Group Identity Formation and Hall's Morphology of Systems all describe idealized processes for the formation of a system and its dynamic relation with the environment. Taken together, we can abstract a generic systems process, in which adaptive systems draw information from their environment, process it, choose a goal, and learn to enact that goal back onto the environment, including the role of recursion and reflection. After establishing the abstract construct, I will explore its usefulness in the realm of education and social change.
BIO: B. Max Grad has a Master's in Teaching from Seattle University and an B.A. in History, Social Studies and Mathematics. After teaching mathematics at the high school level for 6 years,he left the teaching field to pursue a PhD in Systems Science at PSU. His interest is in the nature and structure of learning in individuals, groups and societies.
Here is the link for people who want to attend our seminar remotely:
https://sas.elluminate.com/m.jnlp?sid=2009262&password=M.F9C036DC5C803D8988EBA6BEE53C5A
DATE: Friday, April 13th, 2012, 12:00 - 12:50 PM
PRESENTER: Lee Hullender Rubin, DAOM, LAc
TITLE: Traditional Chinese Medicine's Challenges in Research
SUMMARY: In this seminar, the challenges facing Traditional Chinese Medicine (TCM) researchers will be discussed. A brief overview of TCM's approach to health and wellness, the challenges of researching this whole system of medicine, and a study on placebo acupuncture will be presented. We will discuss the effect of placebo acupuncture on research outcomes. Within this context, we will explore the double-blinded, randomized, controlled trial design role in TCM research.
BIO: Lee Hullender Rubin is a doctor of acupuncture and oriental medicine with more than ten years of clinical experience. She is board certified in acupuncture, Chinese herbs and reproductive medicine. She relocated to Portland, OR, last year to pursue clinical research training at the Oregon Health Science University Human Investigations Program and a post doctoral research fellowship at the Oregon College of Oriental Medicine (OCOM).
Here is the link for people who want to attend our seminar remotely:
https://sas.elluminate.com/m.jnlp?sid=2009262&password=M.5E2B722396DAED802F4991DE2401C3
DATE: Friday, April 20th, 2012, 12:00 - 12:50 PM
PRESENTER: Sergio Antoy
TITLE: Programming with Narrowing
SUMMARY: In this talk, I will introduce narrowing, the characterizing feature of functional logic programming. Narrowing promotes non-determinism and it enables computing with incomplete or unknown information. After a short and informal presentation of Curry, the leading functional logic language, I will discuss a few examples showing that narrowing and its associated non-determinism support very high-level programming.
BIO: Sergio Antoy is a professor of Computer Science at Portland State University, that he joined in 1990, and an associate chair for graduate education. He regularly teaches undergraduate and graduate courses in the area of programming. He received a PhD degree in Computer Science for the University of Maryland in 1987. Sergio's research is on narrowing strategies and the implementation of functional logic programming languages.
Here is the link for people who want to attend our seminar remotely:
https://sas.elluminate.com/m.jnlp?sid=2009262&password=M.6AACDAFE663EF1AC735784AEAA1BFA
DATE: Friday, April 27th, 2012, 12:00 - 12:50 PM
PRESENTER: Stephen Shervais
TITLE: Satisficing vs.Exploring in Multistage Processes
SUMMARY: When is satisficing instead of exploring optimal for managing multistage processes? While normally thought of as an approach for risk averse decision makers with limited resources, satisficing often provides an efficient strategy for applying existing knowledge to complex problems in non-stationary environments. We present a series of simulation studies that demonstrate that a satisficing strategy produces significantly higher payoffs than traditional mixture strategies employed by approximate dynamic programming methods for learning to control multistage processes. We find that satisficing agents have a significantly higher per step payoff than agents that explore the environment over a time horizon sufficient for all agents to learn the optimal policy for the environment. A satisficing agent will take longer to learn the optimal policy, but will complete significantly more tasks in an equivalent time period and thus accumulate a significantly larger reward. Further more, the satisficing agents may learn in parallel without loss of efficiency, so multiple agents may be used to reduce the time to learn the optimal policy. Within the class of strategies characterized by stationary mixtures, satisficing is optimal for managing multistage processes in many uncertain environments.
BIO: Stephen Shervais is an Associate Professor of Management Information Systems at Eastern Washington University.
Here is the link for people who want to attend our seminar remotely:
https://sas.elluminate.com/m.jnlp?sid=2009262&password=M.4B1F7998F91249BDBC79DA98453A30
DATE: Friday, May 4th, 2012, 12:00 - 12:50 PM
LOCATION: Harder House, Room 104
PRESENTER: Bart Massey
TITLE: Understand "Inform 7" as an Ontological Modeling Language
SUMMARY: Graham Nelson's Inform 7 is an ambitious domain-specific programming language for writing Interactive Fiction (IF, aka "text adventure") stories. The most prominent surface feature of Inform is its natural-language syntax. (For example, the title of this talk is a valid Inform sentence.) Less noticeable but more important is the underlying ontology-building language. Inform supports a rich set of concepts including actions, rules, relations, definitions, time, place and state. This conceptual power enables building ontological models that are at once reasonably rich and reasonably manageable. As a result, Inform is more than an IF language: it is a systems description language.
In this talk I will explain the basics of ontology-building and Inform, and argue that the richness of description Inform provides is a requirement for building system ontologies that are worth the trouble.
BIO: Bart Massey is an Associate Professor of Computer Science at Portland State University, with a Ph.D. in Artificial Intelligence from the University of Oregon's Computational Intelligence Research Lab. Bart's previous work in this area includes supervising a M.S. thesis on ontology-building in multiuser text adventure games.
Here is the link for people who want to attend our seminar remotely:
https://sas.elluminate.com/m.jnlp?sid=2009262&password=M.59C7A24ED960787A06509AD2F804CB
DATE: Friday, May 11th, 2012, 12:00 - 12:50 PM
PRESENTER: Shelby Anderson
TITLE: Ceramic Sourcing and Social Networks in Northwest Alaska
SUMMARY: Social networks are essential to human occupation of Arctic environments. Access to non-local goods through networks is also linked to the development of more complex social organization in northern hunter-gatherers groups. Ceramic geochemical and formal data are used to test hypotheses about the nature and extent of networks over the last 1000 years in Northwest Alaska, a period characterized by significant social and environmental change. Results suggest ceramics were circulating more widely than expected and hint at changes in raw material procurement strategies during the study period that may be related to shifts in mobility or networking strategies. Methodological issues associated with applying standard analytical techniques to northern ceramics are also considered with reference to future research.
BIO: Shelby Anderson is an Assistant Professor in the Department of Anthropology at Portland State University. She received her Ph.D. from the University of Washington in 2011. Her research interests include past hunter-gatherer social and technological change, human-environment interactions, ceramic technology, public archaeology and archaeology of the Arctic, Sub-arctic and Pacific Northwest. Shelby also has a long-term interest in contributing long-term archaeological data on human-environment interactions to modern efforts to cope with climate change in the Arctic and beyond. She has pursued these interests through various work and research experiences in Washington, Utah, Colorado, the Russian Far East and Alaska. More info: http://web.pdx.edu/~ashelby/index.html
Here is the link for people who want to attend our seminar remotely:
https://sas.elluminate.com/m.jnlp?sid=2009262&password=M.13D0DD7DD2F39D13FCC4EFB1B520A8
PRESENTER: Mehmet Vurkaç
TITLE: Bayesian and Related Methods: Techniques Based on Bayes' Theorem
SUMMARY: Bayes' theorem is a simple algebraic consequence of conditional probability. Yet, its consequences are critical to philosophy, society, and technology. Starting from its simple derivation, we will show how its interpretation in terms of base rates (priors) and class-conditional likelihoods illuminates everyday problems in medicine and law, and provides signal processing, communications, machine learning, model selection, and other applications of statistics with powerful classification and estimation tools. Next, we will briefly examine some of the ways in which this theorem can be adopted to include multiple attributes, contexts, hypotheses, and levels of risk. Methods derived from or related to Bayes’ theorem include minimax, maximum-a posteriori (MAP), expectation maximization (EM), Markov random fields, hidden Markov models, the Kalman filter, the Viterbi algorithm, and Bayesian Belief Networks.
BIO: Mehmet Vurkaç is currently an Assistant Professor in the Department of Electrical Engineering & Renewable Energy at the Oregon Institute of Technology. He completed his Ph.D. in Electrical & Computer Engineering at Portland State University in 2011. Vurkaç has a B.A. in Math-Physics from Whitman College (1993), and an M.S. in ECE from Portland State University (1999). He has worked in the music industry (Roland Corp.) for four years as a hardware engineer, and served as an adjunct instructor at PSU’s ECE department (2004–09) and at Whitman College in the Music Department (Sound Synthesis, 1994). His dissertation research was in Neural Networks, Reconstructability Analysis, and Computational Ethnomusicology. His current research is in tempo-tracking and onset detection for Afro-Brazilian music in which Eurocentric priors for timing and accent structure do not always hold., Publications:
- On the Need for Clave-Direction Analysis: A New Arena for Educational and Creative Applications of Music Technology. Journal of Music, Technology and Education. Volume 4, Number 1, pp. 27–46, August, 2011.
- A Cross-Cultural Grammar for Temporal Harmony in Afro-Latin Musics: Clave, Partido-Alto and Other Timelines. Current Musicology. Number 94, Accepted, Fall 2012.
https://sas.elluminate.com/m.jnlp?sid=2009262&password=M.43B40A9E6241B3781A08B05BA8ABA4
