• Teaching Individual/Agent-Based Modeling

The Course: Summer Program in Agent-based Modeling

July 27-August 4, 2020 
Humboldt State University, Arcata, California USA

The program will focus on the fundamental early steps of modeling -- formulating and implementing models -- and how to move efficiently through the modeling cycle of testing, analyzing, publishing, and revising models. Participants will be expected to make rapid progress on their own projects during the class but also -- more importantly -- to leave with a clear plan for completion. The program will be based on the strategies and methods described in the instructors’ 2019 textbook.

Participants will be expected to submit the following before the program starts:

  • A preliminary description of their model that identifies (a) the system and problem it addresses, (b) the observations or patterns that will be used to evaluate the model, and (c) the entities and mechanisms represented in it.
  • The model’s current computer implementation, which presumably will be incomplete.

The program’s objectives will then be for each participant to complete the following products by its end:

  • A set of observations or patterns used to design and evaluate their model.
  • A revised and complete description of the model’s purpose, entities, state variables, and scales; and a list of its key mechanisms.
  • A complete outline of the rest of the model description.
  • A set of alternative theories for key agent behaviors.
  • A partial draft of the model software that is tested and documented.
  • A plan for analysis of the completed model to address its purpose.

 

The program will include lectures on essential topics, group discussion and peer review to address basic model design issues by looking at participants’ projects, specialized lectures on topics of interest to participants, and extensive time for independent work facilitated by interaction with instructors and other participants.

The specialized lectures will be selected and scheduled as the program proceeds, to meet specific needs of the participants. These lectures could include topics such as modeling adaptive behavior, model implementation in the NetLogo software platform, integration of models with GIS, existing models of the physical environment, and integration of modeling with field and laboratory research.

 

Preliminary Course Schedule: 2020

Location: TBA

July 27:

  • Introductions
  • Participant presentations summarizing their projects
  • Lecture on the ODD protocol for model description and design
  • Lecture and exercise on pattern-oriented model design
  • Discussion to review participants’ model descriptions and designs

July 28:

  • Independent work on model descriptions and observed patterns for model design
  • Discussion of progress

July 29:

  • Discussion of participants’ software status
  • NetLogo instruction: lecture and exercises
  • Code testing instruction: lecture and exercises

July 30:

  • Independent work on model software
  • Additional NetLogo instruction as needed

July 31:

  • Lecture on theory for adaptive behavior
  • Lecture and exercise on pattern-oriented theory development
  • Independent work with instructor consultation
  • Baseball at the Arcata Ballpark! Humboldt Crabs vs. Chico Nuts

August 1:

  • Lecture on model analysis
  • Exercises on analysis methods
  • Independent work on analysis plans

August 2: Unscheduled

  • Optional excursion to Redwood National Park

August 3:

  • Lecture and discussion on “publication-oriented modeling”: advice and strategies for publishing agent-based research
  • Peer review and independent work on deliverables

August 4:

  • Independent work on presentations
  • The First International Symposium on the Meaning of Agent-based Life: participant presentations on deliverables and what they learned

Optional additions:

  • Further basic NetLogo instruction
  • Focused instruction on NetLogo tools (BehaviorSpace, GIS, Time, ...)
  • Instruction in R and packaged tools for model analysis
  • Guest lecture on the relation between agent-based modeling and empirical (field and lab) research
  • Instruction in model parameterization strategies and tools

Resources