Agent-Based & Hybrid Modeling Bootcamp & Incubator for Health and Health Care: Conceptualization, Formulation, Implementation for Insight

The Agent-Based and Hybrid Modeling Bootcamp & Incubator for Health Researchers is an intensive, hands-on tutorial and dynamic model incubator that seeks to provide systematic, practical and technically detailed guidance in conceptualizing, designing and implementing agent-based and hybrid system science modeling for health and health care, using familiar language and concepts.  This year’s bootcamp & incubator will focus on agent-based modeling, but with a particular emphasis on insightful, practical and performant hybrids with System Dynamics modeling and discrete event modeling, such as for health care service delivery and patient flow.  

To help ease participants in overcoming the learning curve in applying the techniques taught herein to domains of individual interest, those registering for the “incubator” side of the event will further leverage the extensive experience of the instructor, other expert modelers, and teaching assistants (TAs), who will provide ongoing advice, guidance, tips and close hands-on assistance as these participants and their partnered TAs work to build, explore, and debug the start of their own models addressing participant research interests. Participants interested in taking their initial models far further can elect to remain for the full-scale model incubator/hackathon the subsequent week, in which the instructor and teams of TAs will partner with attendees to deliver more evolved models capable of contributing substantive health insights. Attendance to the incubator portion of this event is constrained by the availability of teaching assistants -- please reserve early to ensure an incubator spot.

Using lectures and guided exercises, this bootcamp will introduce a broad set of topics in designing & building agent-based and hybrid modeling, offer optional material for those seeking additional depth of coverage, and emphasize the complementary insights when agent-based and hybrid modeling is combined with traditional health sciences approaches. Course material is drawn from the instructor’s semester-long interdisciplinary courses at MIT and the University of Saskatchewan, as well as previous bootcamps, including those hosted at UCLA, University of New South Wales, NCSU/UNC, Flinders University, Deakin University, the Sax Institute, and beyond.  

Participants will be provided with over 50 example models drawn from across diverse spheres of health and health care.  Health & health care modeling case studies for chronic disease, communicable illnesses, health service delivery, and environmentally driven infectious diseases from around the world will support participants in helping to conceptualize fruitful model designs, scopes and the form of effective modeling projects.  Such models will showcase Agent-Based, System Dynamics and Discrete Event modeling, including -- but not limited to -- use of agents some of whose states are driven by System Dynamics stocks & flows, agent-based epidemiological models linked to discrete event models capturing patient care pathways and patient flow, highly performant hybrid models in which System Dynamics is used for low-risk populations and in which agents are individuated at a certain point in risk progression or where interacting agents drive flow rates, multi-scale models, and case studies combining all three types of modeling.

The optional final day of the bootcamp will give participants the opportunity to brainstorm with the instructor and teaching assistants concerning ideas for modeling projects, and to discuss concrete design and implementation strategies for such projects.