🤖 AI Summary
This work addresses the challenges of anomaly detection in high-dimensional data, where traditional methods often suffer from performance degradation and sensitivity to parameter settings. The authors propose a novel Multidimensional Outlier Detection (MDOD) algorithm that enhances discriminative power by extending the original data with an additional dimension filled with zeros and constructing vectors rooted at a designated observation point. Anomalies are identified through cosine similarity measures between these vectors. By innovatively integrating dimensionality expansion with an observation-point-based mechanism, MDOD significantly improves detection accuracy in high-dimensional settings. Empirical evaluations across multiple datasets demonstrate the method’s efficiency and effectiveness, and the implementation has been made publicly available as the open-source Python package “mdod” on PyPI.
📝 Abstract
We propose a new outlier detection method with multi-dimensional data. The method detect outliers based on vector cosine similarity, with a new dataset built by adding a zero value dimension to original data. When a point in the new dataset is chosen as a measured point, an observation point is constructed as an origin with the only difference in the new dimension having a non-zero value when compared to the measured point. The vector from the observation point to the measured point is then formed, followed by another vector from the observation point to another point in the dataset. We compare the cosine similarity of the vectors to find out abnormal data.