System Dynamics Modeling stands out among modern data science algorithms as a unique method for data analysis because of its emphasis on modeling the interactions and less reliance on having lots of data observations.
Most analytical methods seek to fit parameters that simplify or explain the relationships that account for the observed outcomes. This approach tends to work very well especially when you have a lot of data.
For system dynamics it is slightly different. Instead of relying on a lot of data from observed outcomes you instead model (or describe) the relationships between given states in the system. This is extremely helpful if you have very few observations or collecting them is cost prohibitive.
Modern Uses for System Dynamics Modeling
What is System Dynamics?
System dynamics is the study and modeling of a system.
Academic Origins of System Dynamics
In order to better understand this approach it is good to establish some terminology:
- Stock: Any measure such as quantity at a given location.
- Flow: The speed of change of a measure between two stocks.
- System: The group of all stocks and flows necessary to describe the environment.
- Time: Used for back-testing or fast-forwarding to observe changes in stock based on the modeled interactions.