Self Organizing Maps Outputs | Seeq Knowledge Base

The following outputs are generated by the Self Organizing Maps tool. Depending on your configuration and analysis requirements, you can select one or more outputs from the Outputs section.

Output	Type	Description
Anomalies	Condition	Indicates whether an observation is considered anomalous based on the trained Self Organization Maps model and anomaly detection settings.
Quantization Error	Signal	Measures how closely an observation matches the Self Organization Maps model. Higher values indicate observations that are less similar to the patterns learned during training.
Anomaly Detection Error Threshold	Scalar	The quantization error threshold used to classify observations as anomalous. Observations with quantization errors exceeding this threshold are flagged as anomalies.
Average quantization error	Scalar	The average quantization error calculated from the training data. This value can be used as a reference when evaluating model performance and detecting deviations.
Model input properties	Scalar	Provides information about the model inputs used during training. This output can be used for model validation and troubleshooting.
Quantization (x)	Signal	Quantization values for individual input signals used by the model. The quantization values become visible in the Outputs section after input signals have been selected. Note: The number of Quantization outputs generated depends on the number of input signals configured for the model. One quantization output is created for each input signal. This can help identify which inputs contribute most to changes in the overall quantization error.
Training window	Condition	Identifies the time range used to train the Self Organization Maps model. This output can be used to verify the training period and compare it with scored data.