Centralized Moving Average¶
Calculating the central moving average of the historical sales data is the first step in data cleaning. Three data points on each side of the actual data point are used. If there are any missing data points on any sides, because the data point is near any sides of the time series, then fewer data points are used for the calculation. Deviation is calculated for each data point, too.
A confidence interval for each data point is calculated based on the confidence level the user set in Data Navigator.
Detecting Outlier Data point¶
A data point is detected as an outlier if it is outside of the confidence interval.