Assessing Feature Representations for Instance-Based Cross-Domain Anomaly Detection in Cloud Services Univariate Time Series Data

IoT

Clare Conran¹  John D. Kelleher²  Matthew Nicholson³  Rahul Agrahari³  Haytham Assem⁴ 

[1] ADAPT Research Centre, Dublin City University, D09 RFK0 Dublin, Ireland;ADAPT Research Centre, Technological University Dublin, D07 EWV4 Dublin, Ireland;ADAPT Research Centre, Trinity College Dublin, D02 W272 Dublin, Ireland;Huawei Research, D02 R156 Dublin, Ireland;

关键词: time series analysis;    anomaly detection;    data representation;    cloud monitoring;    AIOPS;   

DOI  :  10.3390/iot3010008

来源: DOAJ

【 摘 要 】

In this paper, we compare and assess the efficacy of a number of time-series instance feature representations for anomaly detection. To assess whether there are statistically significant differences between different feature representations for anomaly detection in a time series, we calculate and compare confidence intervals on the average performance of different feature sets across a number of different model types and cross-domain time-series datasets. Our results indicate that the catch22 time-series feature set augmented with features based on rolling mean and variance performs best on average, and that the difference in performance between this feature set and the next best feature set is statistically significant. Furthermore, our analysis of the features used by the most successful model indicates that features related to mean and variance are the most informative for anomaly detection. We also find that features based on model forecast errors are useful for anomaly detection for some but not all datasets.

【 授权许可】

Unknown