Global Activity Scores
Abstract: We introduce a new global sensitivity measure called global activity score. The new measure is obtained from the global active subspace method, similar to the way the activity score measure is obtained from the active subspace method. We present theoretical results on the relationship between Sobol' sensitivity indices and global activity scores. We present numerical examples where we compare the results of the global sensitivity analysis of some models using Sobol' sensitivity indices, derivative-based sensitivity measures, activity scores, and global activity scores. The numerical results reveal the scenarios when the global activity score has advantages over derivative-based sensitivity measures and activity scores, and when the three measures give similar results.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.