midr

midr-package

<img src="figures/logo.png?package=midr&version=0.5.2" style="float: right" alt="logo" width="120">
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 &amp; Matsumori (2025) [Forthcoming], and the package itself is detailed in Asashiba et al. (2025) tools:::Rd_expr_doi("10.48550/arXiv.2506.08338").

internal

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>.

Ryoichi Asasihba

Learning from Black-Box Models by Maximum Interpretation
Decomposition

Ryoichi Asashiba

Hirokazu Iwasawa

Reiji Kozuma

midr-package function

Maintainer: Ryoichi Asashiba <a href="/link/ryoichi.asashiba%40gmail.com?package=midr&version=0.5.2" data-mini-rdoc="midr::ryoichi.asashiba@gmail.com">ryoichi.asashiba@gmail.com</a>
Authors:<ul>
<li>Hirokazu Iwasawa</li>
</ul>Other contributors:<ul>
<li>Reiji Kozuma [contributor]</li>
</ul>

Author

<img src='figures/logo.png' style='float: right' alt='logo' width='120' />
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 &amp; Matsumori (2025) [Forthcoming], and the package itself is detailed in Asashiba et al. (2025) tools:::Rd_expr_doi("10.48550/arXiv.2506.08338").

midr: Learning from Black-Box Models by Maximum Interpretation Decomposition — midr-package

Maintainer: Ryoichi Asashiba <a href='mailto:ryoichi.asashiba@gmail.com'>ryoichi.asashiba@gmail.com</a>
Authors:<ul>
<li>Hirokazu Iwasawa</li>
</ul>

Other contributors:<ul>
<li>Reiji Kozuma [contributor]</li>
</ul>

midr-package: midr: Learning from Black-Box Models by Maximum Interpretation Decomposition

Description

Arguments

Author

See Also