AURORA: A Unified fRamework fOR Anomaly detection on multivariate time series

Abstract

The ability to accurately and consistently discover anomalies in time series is important in many applications. Fields such as finance (fraud detection), information security (intrusion detection), healthcare, and others all benefit from anomaly detection. Intuitively, anomalies in time series are time points or sequences of time points that deviate from normal behavior characterized by periodic oscillations and long-term trends. For example, the typical activity on e-commerce websites exhibits weekly periodicity and grows steadily before holidays. Similarly, domestic usage of electricity exhibits daily and weekly oscillations combined with long-term season-dependent trends. How can we accurately detect anomalies in such domains while simultaneously learning a model for normal behavior? We propose a robust offline unsupervised framework for anomaly detection in seasonal multivariate time series, called AURORA. A key innovation in our framework is a general background behavior model that unifies periodicity and long-term trends. To this end, we leverage a Ramanujan periodic dictionary and a spline-based dictionary to capture both seasonal and trend patterns. We conduct experiments on both synthetic and real-world datasets and demonstrate the effectiveness of our method. AURORA has significant advantages over existing models for anomaly detection, including high accuracy (AUC of up to 0.98), interpretability of recovered normal behavior ( accuracy in period detection), and the ability to detect both point and contextual anomalies. In addition, AURORA is orders of magnitude faster than baselines.

MFAD: a graph anomaly detection framework based on multi-frequency reconstruction

“…Graph data has attracted significant attention due to its powerful modeling capabilities and scalability, especially in graph anomaly detection, which plays a crucial role in applications such as spam detection, financial fraud detection, and social network analysis [1]. Graph anomaly detection aims to identify patterns (nodes, subgraphs, edges, etc.)…”

Time Series Anomaly Detection for KPIs Based on Correlation Analysis and HMM

“…Some anomaly detection methods based on supervised learning [3,4] can perform fast and accurate anomaly detection by relying on a large number of types of anomaly-labeled data, but they are not suitable for the actual operation and maintenance environment, which contains fewer anomalies. The new unsupervised and semi-supervised learning anomaly detection methods [5][6][7][8] can better adapt to the actual operation and maintenance environment. Some models use RNN and LSTM to analyze time series data, but RNN and LSTM have problems, such as error accumulation and the need for a lot of training memory, which leads to some false positives and false negatives in anomaly detection.…”