Calendar

COMPUTATIONAL SOCIAL SCIENCE FRIDAY SEMINAR

Dale Brearcliffe, MAIS-CSS student
Computational Social Science Program
Department of Computational and Data Sciences
George Mason University

Parallelization of Entity-Based Models in Computational Social Science: A Hardware Perspective

Friday, October 27,3:00 p.m.
Center for Social Complexity Suite
3rd Floor, Research Hall

ABSTRACT:  The use of simulations in exploring theories and hypotheses by social scientists is well documented.  As computer systems have grown in capacity, so have interests by social scientists in executing larger simulations.  Social scientists often approach their simulation design from the top down by selecting an Entity-Based Model (EBM) framework from those that are readily available, thus limiting modeling capability to the chosen framework. Ultimately, the framework is dependent upon what is at the bottom, the hardware that serves as the foundation of the computing system.  One underused hardware architecture supports the simultaneous execution of a problem split into multiple pieces.  Thus, the problem is solved faster in parallel. In this seminar, a selection of parallel hardware architectures is examined with a goal of providing support for EBMs. The hardware’s capability to support parallelization of EBMs is described and contrasted. A simple EBM is tested to illustrate these capabilities and implementation challenges specific to parallel hardware are explored.

COMPUTATIONAL SOCIAL SCIENCE FRIDAY SEMINAR

Nathan M. Palmer, Ph.D. Candidate
Computational Social Science Program
Department of Computational and Data Sciences
George Mason University

A Simple Direct Estimate of Rule-of-Thumb Consumption using the Method of Simulated Quantiles and Cross Validation

Friday, November 3, 3:00 p.m.
Center for Social Complexity Suite
3rd Floor, Research Hall

Campbell and Mankiw (1989, 1990) famously demonstrated that aggregate data supported a model of household consumption in which roughly 50% of agents followed an optimizing strategy while the other 50% followed a “rule of thumb” strategy, consuming their current income. This paper revisits that hypothesis using structural, micro-level, semi-parametric estimation and formally selecting between different models of agent behavior. I find strong evidence supporting a generalization of Campbell and Mankiw (1989, 1990)’s original conclusion: roughly 50% of the population behaves in a way similar to “rule of thumb” consumers, even when the data is allowed to dictate how severe that rule of thumb behavior is. In addition, this paper demonstrates the usefulness and flexibility of both the Method of Simulated Quantiles and K-fold cross validation for selecting between of agent behavior. This type of model selection is crucial for creating agents to populate robust, richly-featured agent-based models of macroprudential and macro-financial systems.

MSQ-Rule-of-Thumb-Consumers-GMU2017

COMPUTATIONAL SOCIAL SCIENCE FRIDAY SEMINAR

Sally Evans, Coordinator
University Dissertation and Thesis Services
Fenwick Library
George Mason University

University Dissertation and Thesis Services:
Here to help you submit your thesis or Dissertation CORRECTLY and ON TIME

University Dissertation and Thesis Services understands that there are many steps in the process toward graduation and it is their goal is to make the process as clear, easy, and stress-free as possible. After Ms. Evans’ presentation, you will have an opportunity to ask questions

There will be no Computational Research and Applications Seminar on Monday, November 20.

Happy Thanksgiving!

COMPUTATIONAL SOCIAL SCIENCE FRIDAY SEMINAR

Keith Waters, Ph.D. Candidate
Schar School of Public Policy and Government
George Mason University

Endogenous Region Formation

COMPUTATIONAL SOCIAL SCIENCE FRIDAY SEMINAR

William B. Rouse, Ph.D.
Alexander Crombie Humphreys Chair
School of Systems and Enterprises
Stevens Institute of Technology

Computational Social Science at Several Levels

Friday, December 8, 3:00 p.m.
Center for Social Complexity Suite
3rd Floor, Research Hall

ABSTRACT: Computational social science can enable understanding – and design – of a wide variety of phenomena at an enormous range of levels. This lecture will address processes, organizations, ecosystems, and society for phenomena associated with disease and medicine, health and well being, and technology adoption, particularly in automobiles. The process level is addressed in terms of scaling and optimization of medical innovations. The level of organizations is considered in the context of health provider corporations’ responses to the Affordable Care Act. The ecosystem level is discussed in addressing population health – integrated delivery of health, education, and social services – in the highly fragmented US ecosystem. The society level is considered in terms of the expected disruptive impacts of driverless cars on automotive, insurance, and finance industries. Approaches to and challenges of modeling at these differing levels are discussed.

COMPUTATIONAL SOCIAL SCIENCE FRIDAY SEMINAR

Niloofar Jebelli, MAIS-CSS Student
George Mason University

Urban Development Through the Lens of Agent-Based Modeling

Friday, December 15, 3:00 p.m.
Center for Social Complexity Suite
3rd Floor, Research Hall

Abstract:   Cities are ever changing and growing phenomenon with many underlying complexities. Through its life cycle, a city experiences various forms of dynamics. Models allow for a better understanding of such complexities and dynamics. The model presented in this talk simulates the dynamics of certain processes such as: an urban market, agent interactions in that market, urban growth, sprawl and shrinkage and gentrification. The purpose of this model is to understand the behavioral pattern of the agents and demonstrate the life cycle of a city based on individual agents’ actions. This model is significant in its integration of various subsystems creating a larger system while observing developers’ behavior. Specifically, the model explores some well-known issues, including the Smith’s rent-gap theory, Burgess’s concentric zones model of urban growth, and Alonso’s bid rent theory. The main results from the model show that the agents move to and reside in properties within their income range, with similar neighbors. This is one of the first models that provides a new lens to explore urban development.

Faisal Hasan, General Manager, Data and Public Policy, CARFAX

Srinidhi Melkote, Head of Business Analytics CARFAX

COLLOQUIUM ON COMPUTATIONAL SCIENCES AND INFORMATICS

Leveling the Playing Field:  Information Asymmetry in the Used Vehicle Buying Process

Monday, January 29, 4:30-5:45
Exploratory Hall, Room 3301

Abstract

In 1970, Economist and Nobel Prize winner (2001) George Akerlof published a study: “The Market for “Lemons”: Quality Uncertainty and the Market Mechanism” (The Quarterly Journal of Economics, Vol. 84, No. 3. (Aug., 1970), pp. 488-500).   In the study, Akerlof attempts to show where, in a market where as seller of a product has more data/information than the buyer of the product, about the product’s quality, will potentially result in “an adverse selection of low-quality products.”   In no place is Akerlof’s theory more represented than in the used vehicle market, where buyers and sellers don’t always have the same information about a vehicle’s quality, potentially resulting in low-quality cars being bought and sold.   In 1986, CARFAX sought to begin to “level the playing field” between buyers and sellers of used vehicles by collecting, analyzing and making relevant data/information available in the marketplace to both buyers and sellers.

• What data does CARFAX collect?
• Why this data vs other data?
• How does the market drive the type of data collected?
• How does CARFAX analyze and make this data available to answer real world questions?

  • Is this car safe?

  • How much should I pay for it?

  • What’s it worth?

  • What’s the risk to insure it?

  • What’s the risk to finance it?

Faisal Hasan is the General Manager, Data & Public Policy at CARFAX.   In his over 17 years at CARFAX, Faisal has been responsible for helping to build CARFAX’s Vehicle History Database through public and private data acquisition efforts across North America, including overcoming legislative and regulatory hurdles to data access. Faisal focuses on CARFAX’s efforts to secure and analyze data to feed the CARFAX “Onetime to Lifetime” Game Plan and develop future CARFAX products. Faisal earned his B.A. in Government & Politics at George Mason University and his M.A. in Government at the Johns Hopkins University. Faisal has been a Fairfax County resident for over 35 years. He is married w/four kids, including a GMU Junior studying Biology and a 2017 GMU Kinesiology graduate.

Sri Melkote is Head of Business Analytics at CARFAX.  Sri has been at CARFAX for 2 years and leads teams responsible for valuation modeling, pricing analytics, marketing research, media measurement and optimization. He has over 12 years of data science experience. Prior to joining CARFAX, Sri developed dynamic pricing systems, personalized recommendation engines and inventory planning systems for the travel industry. He holds a Master’s degree in Mathematics from Purdue University and a Bachelor’s degree in Chemical Engineering from Indian Institute of Technology, Madras. He is the author of several journal articles and a recipient of the 2011 INFORMS Revenue Management and Pricing Practice Award.

COMPUTATIONAL SOCIAL SCIENCE FRIDAY SEMINAR

Professor Robert Axtell
Computational Social Science Program, Department of Computational and Data Sciences, College of Science
Department of Economics, College of Humanities and Social Sciences
Center for Social Complexity
George Mason University

Order without Optimality

Friday, February 2,  3:00 p.m.
Center for Social Complexity Suite
3rd Floor, Research Hall

ABSTRACT:  Long ago Hunter Mill Road in Reston, Virginia, was surrounded by farmland and little travelled. Where it crosses Colvin Run creek a single-lane bridge was adequate for the traffic at the time. However, as the region grew in the post-WWII era the traffic on Hunter Mill Road increased until the single-lane bridge is today a bottleneck for rush-hour traffic, southbound toward the Dulles Toll Road in the morning and northbound in the evening. There are no stop signs on either side of the bridge but warning signs that the bridge is too narrow to handle two vehicles moving in opposite directions at the same time. During periods of low traffic—queue length 0 or 1—the usual first-come-first-served (FCFS) driver discipline is observed at the bridge, by which the first vehicle to approach from either side has the right-of-way. However, during periods of heavy traffic, such as the morning and evening ‘rush hours’, a different behavioral regime is observed: pairs of cars moving in the same direction cross the bridge together, in single file. This departure from the FCFS protocol clearly permits higher vehicular flowrates over the bridge in each direction. However, from the layout of the bridge and vehicular dynamic we demonstrate that the globally optimal behavior for the system during busy times would involve larger numbers of vehicles moving together across the bridge, i.e., groups of 3, 4, or even 5 produce shorter wait times overall, dominating the 2-car norm. We argue that this is an example of a sub- optimal spontaneous order, and go on to suggest that there is little reason to believe that (m)any spontaneous orders operate at anything like peak performance. Essentially, spontaneous orders in this context represent behavioral norms for multi-agent coordination problems that improve on so-called anarchic solutions identified with Nash equilibria of the underlying non-cooperative game. The welfare improvements associated with following the emergent social conventions represent a kind of satisficing solution of the type identified by Herbert Simon as a good description of human behavior in a wide variety of social situations, bounded by cognitive constraints, limited information, and incentive problems. We conclude by arguing that F.A. Hayek implicitly associated his conception of spontaneous order with high welfare, but that this view is untenable. Rather, the more general idea of emergence, which makes no welfare assertions, seems to subsume spontaneous order as a very special case, suggesting that Hayek’s contribution to this area has been largely supplanted by more recent developments in the science of complex systems, as has been argued elsewhere.