Coverage Criterion

Coverage criterion, in various fields, aims to quantify how comprehensively a system or model is tested or trained, ensuring all relevant aspects are considered. Current research focuses on developing novel metrics and algorithms, including those based on graph neural networks, conformal prediction, and combinatorial interaction testing, to achieve more effective and efficient coverage in diverse applications such as automated driving systems, machine learning model robustness, and network optimization. This research is significant because improved coverage criteria lead to more reliable systems and models, impacting areas ranging from autonomous vehicle safety to the trustworthiness of AI in critical applications.