Understanding or controlling a physical system often requires a model of the system, that is, knowledge of the characteristics and structure of the system. A model can be a pre-defined structure or can be determined solely through data. In the case of Kalman Filtering, we create a model and use the model as a framework for learning about the system. This is part of the Control branch of our Autonomous Learning taxonomy from Chapter 1.
What is important about Kalman Filters is that they rigorously account for uncertainty ...
