Typical value

Elastic Stack Elasticsearch
1

In the single metric viewer, I see that the actual value is the value calculated from the actual data, but what about the typical value? How can I determine this typical value? Is the typical value being calculated according to any rules?
Ex:

10:00 record_1: product_1
10:01 record_2: product_2
10:02 record_3: product_3
10:05 record_4: product_4
10:09 record_5: product_5
10:16 record_6: product_1
10:19 record_7: product_7
10:20 record_8: product_2
10:21 record_9: product_1
10:22 record_10: product_1
10:23 record_11: product_1
10:24 record_12: product_1
bucket_span1: 10:00-10:15: product_1 actual=1
bucket_span2: 10:16-10:30: product_1 actual=5
==> Typical value ????
2

Hello @zi_ninja ,

please refer to this paper for details on how typical values are modeled and calculated.

3

Is there a way for me to understand better? I read the link you sent but it's just a general algorithm and I don't clearly understand the terminology when your team applies it to anomaly detection.