
Introduction to Data Streams 27
Figure 2.6: The three aggregation levels in StatStream.
2.3.3 Monitoring Correlations Between Data Streams
Zhu and Shasha (2002) present the StatStream system for monitoring tens
of thousands of high-speed time-series data streams in an online fashion and
making decisions based on them. In addition to single stream statistics such as
average and standard deviation, over sliding windows, StatStream also finds
high correlations among all pairs of streams for a given period of time.
StatStream implements efficient methods for solving these problems based
on Discrete Fourier Transforms and a three-level time interval hierarc ...