Anomaly explanation: A review

Anomaly detection has been studied intensively by the data mining community for several years. As a result, many methods to detect anomalies have emerged, and others are still under development. But during recent years, anomaly detection, just like lot of machine learning tasks, is facing wall. This wall, erected lack trust final users, slowed down usage these algorithms in real-world situations which they designed. Having best empirical accuracy not enough anymore; there need explain their outputs users order increase trust. Consequently, new expression emerged recently: eXplainable Artificial Intelligence (XAI). expression, gathers all that provide explanations output gained popularity, especially with outbreak deep learning. A work devoted literature, but as much explanation. There so on reviews can be found topic. In contrast, we were able find survey explanation particular, while surveys XAI general or neural networks example. With this paper, want comprehensive review field. After brief recall some important algorithms, discovered literature will classified according taxonomy define. stems from an analysis what really when trying anomalies.