January 2019
Beginner to intermediate
670 pages
18h 32m
English
In our example, we were dealing with univariate outliers, which means that we tried to identify outliers along one variable. In multivariate cases, this task can be much more difficult. It is not always enough to look for outliers in each variable separately. Sometimes, you will also need to estimate all included variables at once. A particular data point doesn't have to have stood out on any single one of the variables, but its combination of positions on multiple variables can deem it an anomaly. Detecting multivariate outliers requires different resources, but you should keep in mind that you might encounter cases where you will need to undertake this task.
Read now
Unlock full access