Learning from Black-Box Models by Maximum Interpretation Decomposition
Retrieve Color Theme Information
Color Themes for Graphics
Encoder for Qualitative Variables
Wrapper Prediction Function
Plot MID Breakdowns with ggplot2
Plot MID Conditional Expectations with ggplot2
Plot MID Importance with ggplot2
Plot MID Component Functions with ggplot2
Fit MID Models
Calculate MID Breakdowns
Calculate MID Conditional Expectations
Evaluate Single MID Component Functions
Calculate MID Importance
Plot Multiple MID Component Functions
Extract Terms from MID Models
midr: Learning from Black-Box Models by Maximum Interpretation Decompo...
Encoder for Quantitative Variables
Plot MID Breakdowns
Plot MID Conditional Expectations
Plot MID Importance
Plot MID Component Functions
Predict Method for fitted MID Models
Print MID Models
Color Theme Scales for ggplot2 Graphics
Register Color Themes
Calculate MID-Derived Shapley Values
Summarize MID Models
Default Plotting Themes
Weighted Loss Function
Weighted Sample Quantile
Weighted Tabulation for Vectors
The goal of 'midr' is to provide a model-agnostic method for interpreting and explaining black-box predictive models by creating a globally interpretable surrogate model. The package implements 'Maximum Interpretation Decomposition' (MID), a functional decomposition technique that finds an optimal additive approximation of the original model. This approximation is achieved by minimizing the squared error between the predictions of the black-box model and the surrogate model. The theoretical foundations of MID are described in Iwasawa & Matsumori (2025) [Forthcoming], and the package itself is detailed in Asashiba et al. (2025) <doi:10.48550/arXiv.2506.08338>.
Useful links