News for hackers. Stuff that matters.

A comment on New method identifies the root causes of statistical outliers. Parent comment

Re:New method identifies the root causes of statistical outl... by WastingMyTime89 on Tuesday July 19, 2022 @08:13PM (#32159670)
Obviously you can say that 9999999999 is very different from the rest of the sample consisting of 5 without an implicit distribution. It’s incredibly far from the median compared to all the other value.
[ Parent | on HN ]
Re:New method identifies the root causes of statistical outl... by whatshisface on Wednesday July 20, 2022 @12:22AM (#32161384)
But is it surprising to find it far from the median?
[ Parent | on HN ]
Re:New method identifies the root causes of statistical outl... by WastingMyTime89 on Wednesday July 20, 2022 @12:36AM (#32161464)
It doesn’t matter if it’s surprising. It might be a measurement mistake, a mixed population, a fat tail or a poor sample. It’s different therefore an outlier.
[ Parent | on HN ]
Re:New method identifies the root causes of statistical outl... by nl on Wednesday July 20, 2022 @02:00AM (#32162002)
There are many statistical tests that will tell you that based on the observations alone (with no model). Basically this depends on how many 5s there are.
[ Parent | on HN ]
Re:New method identifies the root causes of statistical outl... by whatshisface on Wednesday July 20, 2022 @10:56AM (#32166644)
Every single one of those tests makes an implicit assumption about the distribution, explicit in its derivation. Usually, it is that the distribution is Gaussian.
[ Parent | on HN ]