Handling outliers in time series is a complex topic, and, any help that NumXL can offer in detecting (potential) outliers in time series data set is highly desirable;
I am looking for a way to identify suspect outliers, and what to do with outliers observations (e.g. substitute with a plug-in value, interpolated value, etc.).
Finally, can we simulate different kind of outliers as part of our ad-hoc simulation scenarios to conduct a stress-testing?