Nowadays, we have been dealing with a large amount of data in which anomalies occur naturally for many reasons, both due to hardware and humans. Therefore, it is necessary to develop efficient tools that are easily adaptable to various data. The paper presents an innovative use of classical statistical tools to detect outliers in multidimensional data sets. The proposed approach uses well-known statistical methods in an innovative way and allows for a high level of efficiency to be achieved using multi-level aggregation. The effectiveness of the proposed innovative method is demonstrated by a series of numerical experiments.

中文翻译：

---

直观的自适应异常值检测器

如今，我们一直在处理大量数据，其中异常自然发生的原因有很多，包括硬件和人为。因此，有必要开发易于适应各种数据的高效工具。本文提出了一种创新地使用经典统计工具来检测多维数据集中的异常值。所提出的方法以创新的方式使用众所周知的统计方法，并允许使用多级聚合来实现高效率。通过一系列数值实验证明了所提出的创新方法的有效性。