Loading...
Outlier detection for temporal data /
Outlier (or anomaly) detection is a very broad field which has been studied in the context of a large number of research areas like statistics, data mining, sensor networks, environmental science, distributed systems, spatio-temporal mining, etc. Initial research in outlier detection focused on time...
Main Authors: | , , , |
---|---|
Format: | eBook |
Language: | English |
Published: |
San Rafael, California (1537 Fourth Street, San Rafael, CA 94901 USA) :
Morgan & Claypool,
2014.
|
Series: | Synthesis digital library of engineering and computer science.
Synthesis lectures on data mining and knowledge discovery ; # 8. |
Subjects: | |
Online Access: | Abstract with links to full text |
Table of Contents:
- 1. Introduction and challenges
- 1.1 Temporal outlier examples
- 1.2 Different facets of temporal outlier analysis
- 1.3 Specific challenges for outlier detection for temporal data
- 1.4 Conclusions and summary
- 2. Outlier detection for time series and data sequences
- 2.1 Outliers in time series databases
- 2.1.1 Direct detection of outlier time series
- 2.1.2 Window-based detection of outlier time series
- 2.1.3 Outlier subsequences in a test time series
- 2.1.4 Outlier points across multiple time series
- 2.2 Outliers within a given time series
- 2.2.1 Points as outliers
- 2.2.2 Subsequences as outliers
- 2.3 Conclusions and summary
- 3. Outlier detection for data streams
- 3.1 Evolving prediction models
- 3.1.1 Online sequential discounting
- 3.1.2 Dynamic cluster maintenance
- 3.1.3 Dynamic Bayesian networks (DBNS)
- 3.2 Distance-based outliers for sliding windows
- 3.2.1 Distance-based global outliers
- 3.2.2 Distance-based local outliers
- 3.3 Outliers in high-dimensional data streams
- 3.4 Detecting aggregate windows of change
- 3.5 Supervised methods for streaming outlier detection
- 3.6 Conclusions and summary
- 4. Outlier detection for distributed data streams
- 4.1 Examples and challenges
- 4.2 Sharing data points
- 4.3 Sharing local outliers and other data points
- 4.4 Sharing model parameters
- 4.5 Sharing local outliers and data distributions
- 4.6 Vertically partitioned distributed data
- 4.7 Conclusions and summary
- 5. Outlier detection for spatio-temporal data
- 5.1 Spatio-temporal outliers (ST-outliers)
- 5.1.1 Density-based outlier detection
- 5.1.2 Outlier detection using spatial scaling
- 5.1.3 Outlier detection using Voronoi diagrams
- 5.2 Spatio-temporal outlier solids
- 5.2.1 Using Kulldorff scan statistic
- 5.2.2 Using image processing
- 5.3 Trajectory outliers
- 5.3.1 Distance between trajectories
- 5.3.2 Direction and density of trajectories
- 5.3.3 Historical similarity
- 5.3.4 Trajectory motifs
- 5.4 Conclusions and summary
- 6. Outlier detection for temporal network data
- 6.1 Outlier graphs from graph time series
- 6.1.1 Weight independent metrics
- 6.1.2 Metrics using edge weights
- 6.1.3 Metrics using vertex weights
- 6.1.4 Scan statistics
- 6.2 Multi-level outlier detection from graph snapshots
- 6.2.1 Elbows, broken correlations, prolonged spikes, and lightweight stars
- 6.2.2 Outlier node pairs
- 6.3 Community-based outlier detection algorithms
- 6.3.1 Community outliers using community change patterns
- 6.3.2 Change detection using minimum description length
- 6.3.3 Community outliers using evolutionary clustering
- 6.4 Online graph outlier detection algorithms
- 6.4.1 Spectral methods
- 6.4.2 Structural outlier detection
- 6.5 Conclusions and summary
- 7. Applications of outlier detection for temporal data
- 7.1 Temporal outliers in environmental sensor data
- 7.2 Temporal outliers in industrial sensor data
- 7.3 Temporal outliers in surveillance and trajectory data
- 7.4 Temporal outliers in computer networks data
- 7.5 Temporal outliers in biological data
- 7.6 Temporal outliers in astronomy data
- 7.7 Temporal outliers in web data
- 7.8 Temporal outliers in information network data
- 7.9 Temporal outliers in economics time series data
- 7.10 Conclusions and summary
- 8. Conclusions and research directions
- Bibliography
- Authors' biographies.