Dakota 6.15 has been released. See the release notes for further details. A few of the highlights include:

Highlight: ML/MF UQ Methods

Significant extensions to multi-level (ML) / multi-fidelity (MF) sampling methods for UQ, including iterated versions of approximate control variate (ACV) and multifidelity Monte Carlo (MFMC) methods as well as refinement of previous multilevel and control variate approaches. These methods leverage the new ensemble surrogate model abstraction deployed in v6.14, which affords specification of model fidelities in an unordered / non-hierarchical manner. 

Enabling / Accessing: New ML / MF sampling methods are specified as approximate_control_variate or multifidelity_sampling in combination with the new non_hierarchical surrogate model. Previous methods based on the hierarchical surrogate model are also now more fine-grained, including multilevel_sampling for multilevel Monte Carlo, multifidelity_sampling again for control variate Monte Carlo (the previous two-model hierarchical case is a special case of non-hierarchical MFMC with M approximation models), and multilevel_multifidelity_sampling for multilevel-multifidelity Monte Carlo.

Documentation: Within the Dakota 6.15 Reference Manual, refer to method specifications and options for approximate_control_variate, multifidelity_sampling, multilevel_sampling, and multilevel_multifidelity_sampling.