1Blurb:: Sample allocation approach for multifidelity expansions
2
3Description::
4
5Multifidelity surrogate approaches, including polynomial chaos,
6stochastic collocation, and function train, can optionally employ a
7integrated greedy competition across the model sequence, where each
8model index can supply one or more refinement candidates which are
9competed to determine the candidate with the greatest impact on the
10QoI statistics per unit cost.  This greedy competition implicitly
11determines the optimal sample allocation across model indices.
12
13<b> Default Behavior </b>
14
15The default, when \c allocation_control is not specified, is to
16compute or adapt separately for each model index (individual instead
17of integrated refinement).
18
19Topics::
20
21Examples::
22
23Theory::
24
25Faq::
26See_Also::
27