A Theory of Interpretable Approximations

Maximilian Thiessen
Marco Bressan
Emmanuel Esposito
Shay Moran
Nicolo Cesa-Bianchi
The Thirty Seventh Annual Conference on Learning Theory, June 30 - July 3, 2023, Edmonton, Canada, PMLR (2024), pp. 648-668

Abstract

We define an interpretable approximation of a function $c$ to be an approximation of $c$ using a concise aggregation of simple functions from a class $H$. For example, \ldots decision trees.
×