Chapter 1: Anomaly Detection in Time Series

1.1 Time series and anomaly detection
1.2 Box plot method
1.3 Normal statistical method
1.4 Distance method
1.5 Extreme values
1.6 Percentage threshold adjustment
1.7 Channel threshold adjustment

Chapter 2: Derived Time Series

2.1 Detecting anomalies with derived time series
2.2 Rate of change
2.3 Dispersion
2.4 Main line
2.5 Fluctuation amplitude
2.6 Fluctuation frequency

Chapter 3: Alarm Intensity for Anomaly Detection

3.1 Anomaly detection and alarm intensity
3.2 Linear decay function
3.3 Exponential decay function
3.4 Logarithmic decay function
3.5 Triangular decay function
3.6 Characteristics of decay functions
3.7 Composite alarm intensity

Chapter 4: Anomaly Detection in Multidimensional Time Series

4.1 Multidimensional time series and anomaly detection
4.2 Multidimensional combination
4.3 Multidimensional derivation
4.4 Multidimensional aggregation
4.5 Spatial dispersion

Chapter 5: Detecting Specifically Shaped Curve Segments

5.1 Shape and trend characteristics
5.2 Feature indices
5.3 Inverse standardization of parameters
5.4 Shape discovery process
5.5 Examples of shape discovery
5.6 Discovering consecutive multi-shape segments
5.7 Discovering unstable segments
5.8 Time series similarity

Chapter 6: Linear Fitting Under Mass Conservation Constraints

6.1 Yield
6.2 Bounded linear fitting
6.3 Calculating boundaries using the error limitation method
6.4 Linear fitting under mass conservation constraints

Chapter 7: Unsupervised Clustering Based on Output Importance

7.1 Output yield and production route
7.2 Initialization of centroids
7.3 Prediction
7.4 Hierarchical k-means clustering