liveBook · Manning

Time series analysis

As explained in Outlier Detection in Python

Time-series analysis is the analysis of observations ordered over time to characterize trends, cycles, dependencies, or unusual temporal behavior.

Time-series analysis examines data ordered over time to identify unusual trends, cycles, or other temporal patterns that row-level checks can miss. section 4.13 section 1.1.1

It complements single-row outlier detection by examining sequences, ordering, frequencies, and gaps. Records may look normal individually while their timing or collective arrangement is unusual. section 4.13

Useful temporal signals include:

  • Unusual trends or cycles in financial records. section 1.1.1
  • Spikes in purchase records tied to a supplier at the end of each month. section 1.1.1
  • Gaps in time with no staff starting, or clusters of staff starting within a few days. section 4.13

How Outlier Detection in Python covers it

How deep it goes
Defines temporal anomaly analysis and examples but defers preprocessing, training, and ongoing monitoring details.
Examples
Uses supplier-linked monthly purchase spikes, clustered start dates, and gaps in staff records.
What you can do after
Diagnose unusual temporal behavior in records as part of broader fraud and outlier investigations.
How it connects
Connects outlier detection with collective outlier detection, text analysis, and network analysis.

Also explained in

Compare how 4 books explain time series analysis →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Outlier Detection in Python. The section links point to the parts of the book it draws on.