Emory University Goizueta Business School Strategic Modeling and Social Dynamics

Transcription

1 Emory University Goizueta Business School Strategic Modeling and Social Dynamics Instructor: MJ Prietula, Professor, GBS 410, BB Messages for the course Course: Mon, Wed (10am 11.15am); GBS Room 421 Hours: Arrange time, BB Messages, Skype (mj.prietula) Though simulation and modeling has been part of both science and practice for a very long time. Every once in a while, something different and valuable develops (i.e., different does not necessarily imply added value). In this course we will examine something different that developed in simulation and modeling agent based systems (ABS). ABS build on such concepts as discrete event techniques and object-oriented programming. They fill a niche or two that other approaches cannot fill; therefore, they are used in complement to other approaches. However, the growing use of them in isolation of more traditional methods speaks to their value. ABS afford some unique characteristics that allow managers and researchers to ask different questions, and to ask questions differently. ABS is extremely useful when some of the heroic assumptions of other methods cannot be met. ABS is also capable of investigating symbolic constructs as well as numerical constructs of interest. In particular, ABS has been very successful in modeling social systems that vary in size from two agents to millions. ABS allows the modeler to look into the details of the events that are generating the phenomena of interest, whether it is the behavior of a business, market, government agency, or a group interacting on the Internet. ABS has been quite useful in determining the various types of consequences occurring in what might be called complex systems. Unforeseen consequences of strategy and policy are daily reading in our news. In this course we will examine the fundamentals of simulation and examine the elements of ABS using a rather useful tool NetLogo. It is a programming course that has no prerequisites. In this course we will autopsy ABS models and build others. In addition, we will also explore another type of simulation framework systems dynamics. You will work in teams (well, except for the two exams). You will learn how to design, build, refine, test, and document ABS. For the course project, you will play the role of a consulting team that will conduct a synthetic experiment (i.e., a simulation) on a topic of your choice. You all must learn the basics of NetLogo so that you can all communicate aspects of the model. However, others may elect to become quite proficient and engage more complex projects. There is an assumption that all of you will master the fundamentals of programming (which will be well described for you). The online reference for NetLogo is good. As we shall see, you do not need complex code to generate complex results. In fact, some of the most interesting models are rather simple. This gets to a key element of the course the most important element in modeling is the modeler. Is the question the model is addressing meritorious? The value of the model, dear Brutus, is not in our code, But in ourselves Ca$$iu$ 1

2 Class Procedures: Professional behavior, of course, is expected and the class will be interactive, thus requiring your presence and informed participation (where discussion is supported and placed in the context of the readings and assignments). Grading: Exams (2) 40% [In class, closed notes] Homework (3) 21% Group Project (1) 25% Final Exam 0% [There is no final exam.] Participation, etc. 14% There are two in-class exams. There will be four home works. There will also be informed student led discussions by students as part of participation grade. Students will be selected (without replacement) to lead these discussions. Participation is based substantive comments derived from the readings and discussions. The project will be explained in class. Note that grading is based on the suggested distribution of electives by the GBS. Thus, in the end it is a comparative performance score. Note also that it is your responsibility to demonstrate your competence, interest, and willingness to participate. I intend to give you every opportunity to demonstrate these. Computer/Phone Policy. Please bring your laptops but do not use laptops during class unless requested to do so by the instructor. Please turn off all audible alarms and ringers on cellular phones. Attendance. Attendance is, of course, required, and so is presence which is demonstrated by informed participation. Bring your Name Cards. Name cards are required for participation. Rules of Presence. You are expected to participate in class discussions. You have a right to your opinion and may express it. You must grant that right to others. Cold Calls The policy for the class is to assume you are prepared and I will call on you to summarize, lead the class on a topic, or generate discussion questions. If you are not prepared, please inform me prior to the start of class. Ethics and Code of Conduct All students are expected to adhere to the GBS Honor Code. 2

3 BlackBoard (BB) You are expected to monitor the class BB site daily for important (and not so important) information, assignments, slide postings, blog participation, examples, response to questions, and so forth. Readings & Other Resources These will be handed out or posted on the course BB site. For the most part, readings will be selected as the course unfolds as they fit to the class. 3

4 Schedule Summary (subject to change as life unfolds) Week Meeting Date Title Topic 1 1 Aug 27 Who am I and why am I here? Course description, introductions Why Simulations? 2 2 Sept 3 Why ABM Simulations? Using ABMs in practice & science 3 3 Sep 8 How to build a model I Designing an ABM project 4 Sep 10 NetLogo I, Examples IDE, objects, structures, worlds 4 5 Sep 15 How to build a model II Types of ABM models I 6 Sep 17 NetLogo II, Examples Coding patterns & style 5 7 Sep 22 How to build a model III Types of ABM models II 8 Sep 24 NetLogo III, Examples Types of ABM models III 6 9 Sep29 Sensitivity, Uncertainty Analysis Here be Dragons! 10 Oct 1 Verification & Validation Getting it right 7 11 Oct 6 TBD Consultation & Review 12 Oct 8 Exam 1 No Notes Oct No meeting: Fall Break (Oct 15-16) 13 Oct 15 Experiments I Theory, Design & Hypotheses 9 14 Oct 20 Experiments II Data & Analysis 15 Oct 22 Project Draft Presentations [practice, feedback] Oct 27 Complexity, Emergence, SWN Simple rule, complex behaviors 17 Oct 29 ABMs & Theory Building Theory in the code? Nov 3 Agent Network Analysis I ABMs building SNs 19 Nov 5 Agent Network Analysis II ABMs building SNs Nov 10 Systems Dynamics I SD Basics 21 Nov 12 Systems Dynamics II Strategic Modeling & SD Nov 17 AI Agents Mind Matters 23 Nov 19 (open class) [depends] Nov 24 (open class) [depends] --- Nov 26 No meeting: Thanksgiving Recess (Nov 26-28) Dec 1 TBD Consultation & Review 26 Dec 3 TBD Consultation & Review Dec 8 Exam 2 No notes. Not cumulative. 4

5 Readings This is the initial set of required core documents/readings/examples. Included are some classics and new research. Additional readings & YouTube videos will be added based on the flow and interests of the class and current events. Readings are to be done prior to class. If you have questions regarding the readings, bring them up in class for discussion. NetLogo Documentation (online, accessible via NetLogo HELP Menu) User Manual & Dictionary NetLogo Models Library (accessible via NetLogo program IDE) Axelrod, R. (2012).Launching The Evolution of Cooperation. Journal of Theoretical Biology, 299, Axelrod, R. (1997). Advancing the art of simulation in the social sciences. Working Paper , Santa Fe Institute, Santa Fe NM. Axtell, R., Axelrod, R., Epstein, J. & Cohen, M. (1996). Aligning simulation models: A case study and results. Computational and Mathematical Organization Theory, 1(2), Burton, R., & Obel, R. (2011). Computational Modeling for What-Is, What-Might-Be, and What-Should- Be Studies And Triangulation. Organization Science, 22(5), Burton, R. & Obel, R. (1995). The validity of computational models in organization science: From model realism to purpose of the model. Computational and Mathematical Organization Theory, 1(1), Carley, K. (2002). Computational organization science: A new frontier. Proceedings of the National Academies of Science, 99, Epstein, J. (2008). Why model? Journal of Artificial Societies and Social Simulation vol. 11, no Available at: Epstein, J. (2009). Modeling to contain pandemics. Nature, 460, 687. Farmer, J. & Foley, D. (2009). The economy needs agent-based modeling. Nature, 460, Gode, D. & Sundar, S. (1993). Allocative efficiency of markets with zero-intelligent traders: Market as a partial substitute for individual rationality. Journal of Political Economy, 101(1), Kornhauser, D., Wilensky, U. & Rand, W. (2009). Design guidelines for agent based model visualization. Journal of Artificial Societies and Social Simulation vol. 12, no. 2. Available at: Levine, S. & Prietula, M. (2013). Open collaboration for innovation: Principles and performance. Organization Science. Articles in advance: 30 December. Macal, C., North, M. et al. (2014). Modeling the transmission of community associated methicillinresistant Staphylococcus aureus: A dynamic agent-based simulation. Journal of Translational Medicine, 12: 124. Available at: Prietula, M. (2011). Thoughts on complexity and computational models. In P. Allen, S. Maguire & B. McKelvey, The SAGE Handbook of Complexity and Management (pp ). Thousand Oaks, CA: Sage. Rahmandad, H. & Sterman, J. (2012). Reporting guidelines for simulation-based research in social sciences. Systems Dynamics Review, 28(4), Riolo, R., Cohen, M. & Axelrod, R. (2001). Evolution of cooperation without reciprocity. Nature, 414, R.P. (1992). The third branch of science debuts. Science, 3 April, Sterman, J. (2002). All models are wrong: Reflections on becoming a systems scientist. Systems Dynamics Review, 18, Watts, D. & Strogatz, S. (1998). Collective dynamics of small-world networks. Nature, 393, Weinberger, S. (2011). Web of War. Nature, 471,