search_model_graphs

Perform a recursion of admissible DAGs for a given model ensemble

Specification

  • Alias: None

  • Arguments: None

  • Default: NO_GRAPH_RECURSION

Child Keywords:

Required/Optional

Description of Group

Dakota Keyword

Dakota Keyword Description

Optional

model_selection

Perform a recursion of admissible model subsets for a given model ensemble

Required (Choose One)

DAG Ensemble Generation Option

no_recursion

Do not recur over admissible DAGs for a given model ensemble

kl_recursion

Model graph enumeration that follows the ACV-KL partitioning scheme

partial_recursion

Perform a partial recursion of admissible DAGs

full_recursion

Perform a full recursion of all admissible DAGs for a given model ensemble

Description

Within the context of generalized ACV ([BLWL22]), search over a set of admissible directed acyclic graphs (DAGs) for a given model ensemble. The DAG with the best performance (lowest estimator variance for a prescribed budget or lowest cost for a prescribed accuracy) is selected for final post-processing.

In order of increasing numbers of DAGs to enumate, options include kl_recursion, partial_recursion with a depth_limit, and full_recursion.

Theory

Refer to [BLWL22] for additional details on generalized ACV recursion.