A linear model is an extremely basic form of statistical machine learning.
This code strongly based off of Jeremy Kun's post about linear models. However, after looking around, it seems this method is actually pretty universal.
I'd might eventually extend this to be more useful.
linear-model.rkt
provide
s a function get-best-fit
. You hand it a list of cons y-val x-vals
and it returns the coefficients of the function the funciton itself. For an example, see housing.rkt
.