The Sparse Grid Interpolation Toolbox is a Matlab
toolbox for recovering (approximating) expensive,
possibly high-dimensional multivariate functions. It
includes hierarchical sparse grid interpolation
algorithms based on both piecewise multilinear and
polynomial basis functions. Sparse grids are
superior to conventional (full) tensor-product grids
due to a significant reduction of the support nodes.
The asymptotic error decay of full grid interpolation
is preserved up to a logarithmic factor provided
that the objective function is smooth enough.
The toolbox also includes efficient dimension-
adaptive algorithms that automatically detect full
or partial separability of the objective model,
thereby performing well even for large problem
dimensions d > 10 (up to several hundreds,
depending on the
Treatment of models with multiple
output parameters (possibly several hundreds) is
Since version 3.5, accurate gradients can be
computed at very low additional cost.
Since version 4.0, efficient algorithms are provided
to search the interpolant for minima/maxima.
Since version 5.0, numerical integration using
sparse grids is supported (including Gauss-
Patterson sparse grid).