The article entitled "Corner Cases in Machine Learning Processes" by Florian Heidecker, Maarten Bieshaar and Bernhard Sick was published in the journal AI Perspectives & Advances. In the field of machine learning (ML)-based applications, such as highly autonomous driving, the performance of the ML model is of paramount importance. This thesis deals with the definition of corner cases from the perspective of an ML model, addressing aspects that are crucial for understanding rare and potentially dangerous situations. The thesis gives an overview of the characteristics of corner cases and provides a comprehensive description and mathematical formulation, focusing on relevance-weighted loss. By operationalizing these characteristics, the study contributes to an extended taxonomy for ML corner cases that incorporates input, model and application levels to provide a nuanced understanding of corner case properties.