An Introduction to Multi-series Anomaly Detection
Kevin Feasel · Apress eBooks · 2022
Chapters 13 and 14 dealt with one form of time series anomaly detection: the single time series. With that understanding in place, we will extend our focus to situations in which there are multiple series of data that we want to analyze in conjunction rather than independently. In this chapter, we will formalize our understanding of multi-series time series, focusing on how it differs from the single series approach we took earlier. Then, we will look at four common techniques for analyzing multi-series data. Before we wrap up this chapter, we will review some common problems when trying to analyze multi-series data before we implement one algorithm in Chapter 16 .