Research Seminar in System Dynamics
Doctoral seminar in system dynamics modeling, with a focus on building advanced modeling and research skills. Topics vary from year to year and may include: classic works in dynamic modeling from various disciplines (e.g., psychology, sociology, behavioral economics) and current research problems and papers; advanced system dynamics models focused on research and practical problems of interest to students; analytic tools and methods for model development, estimation, and analysis (e.g., automating modeling workflow, maximum likelihood, simulated method of moments, dynamical games, dynamic programming); bootcamp for enhancing modeling skills working on multiple problem sets.
Spring 2025 · Building Serious System Dynamics Models from Scratch
The focus of this advanced research seminar is on building robust system dynamics models from scratch. This class helps students build hands-on modeling skills that allow them to tackle their own modeling projects independently.
Spring 2014 · Advanced Research Seminar in System Dynamics:
Topics
- Model Analysis Process; Equilibrium
- Model Replication and Analysis
- Formulations: Allocation, Coflows
- Useful functional forms, Subscripts, Data Connections
- Process Noise, Sensitivity and Monte Carlo Simulation
- Tutorial Submission due
- Optimization (deterministic and stochastic)
- Calibration (Maximum Likelihood and simulated method of moments)
- Calibration (Kalman Filtering and confidence intervals)
- Dynamic Control and Games
- Advanced Model Analysis Techniques
- Modeling that Matters; Final notes
Readings
- Sterman, John D. Business Dynamics: Systems Thinking and Modeling for a Complex World . McGraw-Hill Higher Education, 2000. ISBN: 9780072311358.
- [BD]= Sterman, John D. Business Dynamics: Systems Thinking and Modeling for a Complex World . McGraw-Hill Higher Education, 2000. ISBN: 9780072311358.
- [BD] Chapter 3.
- Wolstenholme, Eric F. “A Systematic Approach to Model Creation.” In Modeling for Learning Organizations . Edited by John D. W. Morecroft and John D. Sterman. Productivity Press, 1994. ISBN: 9781563270604.
- Skim: Shantzis, Steven B., and William W. Behrens III. “Population Control Mechanisms in a Primitive Agricultural Society.” System Dynamics Group, Alfred P. Sloan School of Management, M.I.T. 1972.
- Moxnes, Erling. “Optimization and System Dynamics as Complements.” (2014).
- Struben, Jeroen, John Sterman, and David Keith. “Parameter and Confidence Interval Estimation in Dynamic Models: Maximum Likelihood and Bootstrapping Methods.” Working Paper, 2014.
- Eberlein, Robert L. “Working with Data using Filtering and State Resetting.” Working Paper, 2014.
- Osgood, Nathaniel, et al. “An Introduction to Using Markov Chain Monte Carlo Approaches with Dynamic Models.” Working Paper, 2014.
- Osgood, Nathaniel, and JuxinLiu. “Towards Closed Loop Modeling: Evaluation the Prospects for Creating Recurrently Regrounded Aggregate Simulation Models Using Particle Filtering.” Proceedings of the 2014 Winter Simulation Conference . Edited by A. Tolk, S. D. Diallo, I. O. Ryzhov, L. Yilmaz, S. Buckley, and J. A. Miller, 2014.
- Bertsekas, Dimitri P., and John N. Tsitsiklis. “Introduction.” In Neuro–Dynamic Programming . Athena Scientific, 1996, pp. 2–10. ISBN: 9781886529106.
- Rahmandad and Spiteri. “An Introduction to Differential Games for Dynamic Modelers.” Working Paper, 2014.
- Andersen, Edward G., and Nitin R. Joglekar. “Using Optimal Control Theory with Dynamic Models.” Working Paper, 2014.
- Oliva, Rogelio. “Eigenvalue Elasticity Analysis.” (2014). Draft.
Prerequisites & course information
Prereq: 15.873 and permission of instructor G (Fall, Spring) 3-0-9 units Can be repeated for credit.
Sources
- Course catalogue ↗2026–2027
- MIT OpenCourseWare ↗Spring 2025
- MIT OCW syllabus ↗
- MIT OpenCourseWare ↗Spring 2014
- MIT OCW syllabus ↗
- MIT OCW calendar ↗
- MIT OCW readings ↗
- Historical MIT OpenCourseWare offering. Named reading entries and schedule topics are automatically extracted from the linked source pages.
- Historical MIT OpenCourseWare offering. Named reading entries and schedule topics are automatically extracted from the linked source pages.