Over the last years, artificial intelligence (AI) models have become so complex that understanding them has raised the question about their interpretability. The terms interpretability and explainability have been used by researchers interchangeably. These two terms sound like very closely related, but according to some works one has to distinguish these two concepts. Interpretability is mostly related to the outcome of the cause and-effect relationship given the system's inputs. Explainability deals with the internal logic of a machine learning system. The aim is to characterize model accuracy and transparency in AI-powered decision making. It is clear that there is a need for a proper mathematical formalism that is still missing. Hence, there is a trade-off between the performance of a machine learning model and its ability to produce explainable and interpretable predictions. The study of robust systems which are also explainable and interpretable is still under way. Explainability and interpretability have become a requirement to comply with government regulations for sensitive applications, such as in finance, public health, and transportation. In fact, this issue has received attention from the European Parliament whose General Data. Protection Regulation recognizes the right to receive an explanation for algorithmic decisions. This also justifies the attention on this topic.This Special Issue aims to collect some advancements in the field. The details of the special issue can be found from the. link https://www.springer.com/journal/12065/updates/19332152.------------------------------
Yichuan Zhao
Georgia State University
------------------------------