Jonathan Jakob, Martina Hasenjäger, Barbara Hammer, "Reject Options for Incremental Regression Scenarios", International Conference on Artificial Neural Networks (ICANN) 2022, 2022.

Machine Learning with a Reject Option is the empowerment of an algorithm to abstain from prediction when the outcome is likely to be inaccurate. Although, already studied many decades ago, this field of machine learning has recently gained some traction again. However, most reject option applications concern themselves with classification tasks and from the little work that is available for regression systems all are about rejections in an offline setting. In this publication, we study the problem of reject options for online regression. We compare and evaluate different approaches to facilitate this problem, both in a theoretical and a real world setting, and reach a clear outcome.