Skip to content


Below is a table with all the available outputs from TIM Anomaly Detection.

Configuration field model building model rebuilding detection
Simple importances* x x
Extended importances* x x
Data offsets* x x
Anomaly indicator x x x
Normal behavior x x x
Residuals x x x
Sensitivity x x
Result explanations* x x x

* - defined in the Outputs section inside TIM Forecasting

Anomaly indicator

A number from the interval (0, infinity) that specifies the extent to which a given data point in time is anomalous. Data points with anomaly indicator higher than 1 are considered as anomalies. See this section to learn more.

"anomalyIndicator": {
    "values": {
      "2018-10-30T00:00:00Z": 0.7,
      "2018-10-31T00:00:00Z": 0.1,
      "2018-11-01T00:00:00Z": 1.3

Normal behavior

A real number returned for each data point where the normal behavior behavior learner is evaluated. It describes how the KPI should behave under the circumstances given by explanatory variables.

"normalBehavior": {
    "values": {
        "2018-10-30T00:00:00Z": 15963.314,
        "2018-10-31T00:00:00Z": 15261.849,
        "2018-11-01T00:00:00Z": 15753.487


A real number returned for each data point where the normal behavior learner is evaluated. The number describes the departure from the normal behavior. The residuals time serie constitute the main input of the anomalous behavior learner; it is the basis for calculating other detection features.

"residuals": {
    "values": {
      "2020-05-13T06:00:00.000Z": -1.260717,
      "2020-05-13T06:10:00.000Z": -1.855001,
      "2020-05-13T06:20:00.000Z": 0.250813,


A sensitivity parameter that was used to build the model. If a concrete input sensitivity parameter is specified then it is equal to the output sensitivity. If, however, only the maximum sensitivity parameter is specified then TIM determines the sensitivity automatically and the results is returned in the output sensitivity.

"sensitivity": 0.5