Relationship of different algorithms
I just realized today that proportional derivative control is basically the same as a perceptron algorithm in the application I’m using. Using the PD controller basically allows you to control the relationship of your weights and inputs more, since the perceptron has an associated function that is a pure linear relation between the vector of input vectors and the output vector using the vector of weights.
I haven’t quite thought through everything to see if this corresponds to a reduction of sorts to bring the problem down to linear, but it’s an interesting thing to connect for understanding purposes.

















