fmeffects: An R Package for Forward Marginal Effects

TL;DR

The paper introduces the R package fmeffects, which implements forward marginal effects, a model-agnostic interpretation method for understanding how feature changes impact predictions in complex models.

Contribution

It provides the first software implementation of forward marginal effects theory, enabling practical application and further development of this interpretability approach.

Findings

First software implementation of forward marginal effects

Supports interpretation of non-linear and non-parametric models

Facilitates model explanation through feature step changes

Abstract

Forward marginal effects have recently been introduced as a versatile and effective model-agnostic interpretation method particularly suited for non-linear and non-parametric prediction models. They provide comprehensible model explanations of the form: if we change feature values by a pre-specified step size, what is the change in the predicted outcome? We present the R package fmeffects, the first software implementation of the theory surrounding forward marginal effects. The relevant theoretical background, package functionality and handling, as well as the software design and options for future extensions are discussed in this paper.