Emphasizing prescriptive insights for explainability
Prescriptive insight is a popular jargon used in data analysis. It means providing actionable recommendations derived from the dataset to achieve the desired outcome. It is often considered to be a catalyst in the entire process of data-driven decision-making. In the context of XAI, explanation methods such as counterfactual examples, data-centric XAI, and what-if analysis are prominently used for providing actionable suggestions to the user.
Along with counterfactuals, the concept of actionable recourse in ML is also used for generating prescriptive insights. Actionable recourse is the ability of a user to alter the prediction of an ML model by modifying the features that are actionable. But how is it different from counterfactuals? Actionable recourse can be considered to be an extension of the idea of counterfactual examples, which uses actionable features instead of all the features present in the dataset.
Now, what do...