Goal-Oriented Shapley Effects with Special Attention to the Quantile-Oriented Case

Abstract

Abstract. We propose to study quantile-oriented sensitivity analysis (QOSA) indices and show some of their theoretical properties. These have a number of shortcomings when dealing with both independent and dependent inputs, which leads us to define new generic indices based on the Shapley values named goal-oriented Shapley effects (GOSE). In particular, we focus on quantile-oriented Shapley effects (QOSE) and subsequently perform several calculations of QOSA indices and QOSE in order to better understand the behavior and the respective interest of each.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 5 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.