PRIM for bump-hunting for high-dimensional regression-type data.
Tarn Duong <tarn.duong@gmail.com>
The data are \((\bold{X}_1, Y_1), \dots, (\bold{X}_n, Y_n)\) where \(\bold{X}_i\) is d-dimensional and \(Y_i\) is a scalar response. We wish to find the modal (and/or anti-modal) regions in the conditional expectation \( m(\bold{x}) = \bold{E} (Y | \bold{x}).\)
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 PRIM
estimation for 2- and 5-dimensional data.
Friedman, J.H. & Fisher, N.I. (1999) Bump-hunting for high dimensional data, Statistics and Computing, 9, 123--143.
Hyndman, R.J. Computing and graphing highest density regions. American Statistician, 50, 120--126.