Does anyone have experience working with “anomaly detection” on Dundas
Hi @ben.taicher - Dundas BI version 7.0 is adding a new formula called OutlierInterquartile (called OUTIQ) which should do what you are asking. This formula is based on this concept:
I haven’t yet had a chance to play with this as the official release isn’t for a day or so but hopefully you can check it out soon.
Thanks a lot!
There are two variations for the Outliers. One is using Interquartile computations, the other is using the Tau algorithm. There are further settings that the user can customize.
You might enjoy this new video I created on the outliers formulas. There are two formulas but this will give you an idea of how to use either.