In many cases there’s sequential data (often chronological) that’s supposed to have a fairly steady value or increasing/decreasing at a fairly steady rate. If there are spots in the data where it is not fairly steady, you want to know about it. Or in other words, if you graphically represent the data, you want to find the abnormal peaks and spikes.
As a database professional, an obvious case of this situation is tablespace storage usage. Normally the number of GBs grows approximately the same rate each day/week/month – any excessive growth rate is indicative of an abnormal workload, ...
