Global Outlier

Global outlier detection aims to identify data points significantly deviating from the typical data distribution, a crucial task across diverse fields. Current research emphasizes developing robust and efficient algorithms, including those based on clustering, generative models (like Isolation Forest and BiGANs), and optimal transport, often focusing on handling high-dimensional data and improving interpretability. These advancements are vital for enhancing the reliability of machine learning models, improving data quality in various applications (e.g., recommender systems, medical imaging, anomaly detection in networks), and facilitating more trustworthy insights from complex datasets.