Description Details Author(s) References Examples
MiniMiniMaxUQ seeks to answer the following: How accurately can a black-box function f be emulated from a given set of data? How many evaluations of f are required to guarantee that f can be emulated to a given level of accuracy?
Package: | MiniMiniMaxUQ |
Type: | Package |
Version: | 1.0 |
Date: | 2014-10-20 |
License: | GPL (>= 2) |
Jeffrey Regier and Philip Stark
Maintainer: Jeffrey Regier jeff@@stat.berkeley.edu
Jeffrey Regier and Philip Stark. "Mini-Minimax Uncertainty Quantification for Emulators." arXiv preprint arXiv:1303.3079 (2013).
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 | library(MiniMiniMaxUQ)
X = expand.grid(1:9, 1:9) / 10
f <- function(x) sin(x[1]) + cos(x[2])
f.X = apply(X, 1, f)
K.hat = find.K.hat(X, f.X)
pointwise.uncertainty(X, f.X, K.hat, c(.55,.33))
lower.bound.computational.burden(X, f.X, K.hat, epsilon=.1)
corners.uncertainty.bound(X, f.X, K.hat)
lower.bound.max.uncertainty(X, f.X, K.hat)
upper.bound.max.uncertainty(X, f.X, K.hat)
branch.and.bound.max.uncertainty(X, f.X, K.hat)
uncertainty.confidence.bounds(X, f.X, K.hat)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.