Simply put, it is an "online" model that which means that it learns on the fly. Specifically, it is the optimal online learning algorithm for linear systems wit

Simply put, it is an "online" model that which means that it learns on the fly. Specifically, it is the optimal online learning algorithm for linear systems with Gaussian noise. In a way it is like a primitive RNN, it has internal state, inputs and outputs.