The data are (X1, Y1), …, (Xn, Yn) where Xi is d-dimensional and Yi is a scalar response variable. We are interested in the conditional expectation function m(x) = E (Y | X = x). PRIM is a bump hunting technique introduced by Friedman & Fisher (1999), taken from data mining. PRIM estimates are a sequence of nested hyper-rectangles (boxes).
For an overview of this package, see vignette("prim")
for bump hunting and estimating highest density difference regions.
Install from CRAN:
install.packages("prim")