Post-hoc explanation methods for machine learning models have been widely used to support decision-making. One of the popular is Counterfactual Explanation (CE), also known as Actionable Recourse, which provides a user with perturbation vector features that alters prediction result. Given vector, can interpret it an "action" obtaining one's desired decision In practice, however, showing only of...

Abstract We consider situations where a model for an ordered categorical response variable is deemed necessary. Standard models may not be suited to perform this analysis, being that the marginal probability effects large extent are predetermined by rigid parametric structure. propose use rank likelihood approach in non Gaussian framework and show how additional flexibility can gained modeling ...

After model selection, subsequent inference in statistical models tends to be overconfident if selection is not accounted for. One possible solution address this problem selective inference, which constitutes a post-selection framework and yields valid statements by conditioning on the event. Existing work is, however, directly applicable additive linear mixed models. A novel extension recent c...