Paper ID: 2404.08747

Observation-specific explanations through scattered data approximation

Valentina Ghidini, Michael Multerer, Jacopo Quizi, Rohan Sen

This work introduces the definition of observation-specific explanations to assign a score to each data point proportional to its importance in the definition of the prediction process. Such explanations involve the identification of the most influential observations for the black-box model of interest. The proposed method involves estimating these explanations by constructing a surrogate model through scattered data approximation utilizing the orthogonal matching pursuit algorithm. The proposed approach is validated on both simulated and real-world datasets.

Submitted: Apr 12, 2024