I have a rather technical question. I'm trying to find the Yule-Nielsen factor
n in the equation
I've read that
n is empirically derived from the "best fit" of the model to the training data:
My understanding is that the training data consists of random colour patches and the spectra of each is measured using a spectrophotometer. But how do you build the best-fit model of a collection of spectra? Suppose each spectrum is formulated as a
M by one matrix. How do you come up with a best-fit model without knowing the value of
n? Is the best-fit an
M dimensional plane?