Syllabus

Course Meeting Times

Lectures: 1 session / week, 3 hours / session

Prerequisites

An introductory system dynamics course, e.g. 15.871 Introduction to System Dynamics and 15.872 System Dynamics II (or equivalents) is a prerequisite. Listeners are welcome based on room capacity, but the real benefits of the course come from taking part in projects and doing the readings.

Course Description and Objectives

In this course, we introduce several technical skills for building robust models of dynamic socio-technical systems. We discuss tools and skills for sensitivity analysis and Monte-Carlo simulation, optimization, model calibration, and data manipulation in dynamic models. Through related research projects, students enhance and apply their system dynamics modeling skills to problems of their interest. The skills developed in this course build upon those introduced in 15.871 and 15.872 and are essential for students who want to use their modeling skills on real research and application problems. Moreover, projects in this course provide opportunities for graduate students to find interesting questions for their research program.

Major Course Concepts and Techniques

1. Modeling and model analysis process
2. Model formulation and critique
3. Noise and sensitivity analysis
4. Optimization, model calibration and parameter estimation
5. Optimal control and differential games
6. Advanced modeling and analysis tools

Required Texts

Sterman, John D. Business Dynamics: Systems Thinking and Modeling for a Complex World. McGraw-Hill Higher Education, 2000. ISBN: 9780072311358.

Software

Vensim DSS (Decision Support System). Vensim DSS is Ventana's top-of-the-line version. See the Vensim website for download options.

Hardware

Many class exercises will require the use of a computer.

Reading Assignments

We flip the classroom, so class work relies on the assumption that you have done the readings and we can build on that basic knowledge. The reading assignments will consist of relevant articles, tutorials, and business dynamics and Vensim manual chapters. We assume a general familiarity with Sterman's text, which is a reference book for the course, but additional required reading will be assigned from that. See the Readings section for a list of the assigned readings and additional references.

Weekly Presentations

Each team will be presenting their progress in 20-minute presentations in class every other week, and we will discuss these in the class and provide feedback. Presentations are focused on reporting progress, sharing challenges and problems, and receiving constructive feedback. Simplicity and focusing on what adds value for your project progress is the key for good presentations, so do not spend time on fancy graphics.

Class Exercises

The readings will prepare you to apply the advanced techniques in simple problems. We typically spend the class time beyond project presentations addressing your questions on different methods we cover and working on a few exercises to apply those methods. Your active participation in these activities contributes to your classroom participation component of the grade.

Projects and Tutorials

The course is organized around your team projects, which are expected to be at a quality sufficient to generate journal publications. You select your team and your project by the first session. Teams of 2–4 students will work on the projects. Teams will follow a structured process for enhancing the project. Adherence to this process is critical to enable timely management of the project. You will have presentations of your project progress every other week, providing an opportunity to receive feedback from the instructor and your classmates.

The final report from the project is a manuscript that discusses the problem, the literature, the modeling and analysis work, and your results and insights.

Your team is expected to develop a video tutorial that either introduces one of the methods we cover in the course using a simple example, or clearly shows how to use a specific capability within Vensim to solve a common modeling problem.

Grading