Events

[ERGO] Seminar by Jasone Ramírez-Ayerbe

Description: Title: Novel mathematical optimization models to generate group counterfactual explanationsAbstract: Machine Learning models are increasingly being used in high-stakes decision-making environments such as healthcare, finance or law. Many of these models are black-boxes and therefore do not explain how they arrive at decisions in a way that humans can understand. Counterfactual analysis has proven to be a powerful tool in the growing field of Explainable Artificial Intelligence. In Supervised Classification, the goal is to associate with each record a counterfactual explanation: an instance that is close - according to a given metric- to the record and whose probability of being classified in the positive class (a “good class”) by a given classifier is high. Finding counterfactual explanations is equivalent to solving an optimization model, the structure of which will depend on several ingredients. This talk will illustrate several such models, including when dealing with group of instances, functional data, or when applied to benchmarking models.