HOLT_WINTERS

Compares current value to the value predicted by the Holt-Winters statistic forecasting algorithm¹.

Inputs

"targetProperty": "current": The data on which to perform detection. It should contain the historical data to use for training.

Parameters

name	description	default value
component.sensitivity	Detection sensitivity scale from 0 - 10, mapping z-score from 1 to 3.	5
component.lookbackPeriod	Historical period to use for training. In ISO-8601 format. Requires component.monitoringGranularity, see shared parameters. Eg: P14D. If component.lookbackPeriod is not set, component.lookback is used.
component.lookback	Deprecated. Prefer component.lookbackPeriod. Number of data points to use for training.	60
component.pattern	Detect as an anomaly if the metric drop, rise or both directions. UP, DOWN, UP_OR_DOWN.	UP_OR_DOWN
component.period	Deprecated. Prefer component.seasonalityPeriod. Seasonality biggest period in number of data points.	7
component.seasonalityPeriod	Seasonality biggest period to learn. In ISO-8601 format. Requires component.monitoringGranularity, see shared parameters. Eg: P7D will learn weekly and smaller seasonalities. If component.seasonalityPeriod is not set, component.period is used.
component.alpha	Level smoothing factor. In [0,1]	-1 - Optimized by BOBYQA
component.beta	Trend smoothing factor. In [0,1]	-1 - Optimized by BOBYQA
component.gamma	Seasonal smoothing factor. In [0,1]	-1 - Optimized by BOBYQA

Example

Recommended defaults

{
  "name": "root",
  "type": "AnomalyDetector",
  "params": {
    "type": "HOLT_WINTERS",
    "component.sensitivity": "4",  
    "component.monitoringGranularity": "P1D",
    "component.lookbackPeriod": "P14D",
    ...  # shared parameters
  },
  "inputs": [
    {    # data with historical data for training
      "targetProperty": "current",
      "sourcePlanNode": "currentData",
      "sourceProperty": "currentOutput"
    }
  ]
}

---

1. See https://otexts.com/fpp2/holt-winters.html↩