Model-Agnostic Interpretations with Forward Marginal Effects
Computes AMEs for every feature (or a subset of features) of a model.
R6 Class computing Average Marginal Effects (AME) based on Forward Mar...
Computes a partitioning for a ForwardMarginalEffect
Computes FMEs.
fmeffects
R6 Class representing a forward marginal effect (FME)
User-friendly function to create a Predictor .
R6 Class representing a partitioning
PartitioningCtree
PartitioningRpart
Plots an ForwardMarginalEffect object.
Plots an FME Partitioning.
R6 Class representing a predictor
PredictorCaret
PredictorLM
PredictorMLR3
PredictorParsnip
Prints an ForwardMarginalEffect object.
Prints an FME Partitioning.
Prints summary of an AverageMarginalEffects object.
Prints summary of an ForwardMarginalEffect object.
Prints summary of an FME Partitioning.
Create local, regional, and global explanations for any machine learning model with forward marginal effects. You provide a model and data, and 'fmeffects' computes feature effects. The package is based on the theory in: C. A. Scholbeck, G. Casalicchio, C. Molnar, B. Bischl, and C. Heumann (2022) <doi:10.48550/arXiv.2201.08837>.
Useful links