-
Notifications
You must be signed in to change notification settings - Fork 101
High cardinality #332
High cardinality #332
Conversation
sean-zheng-amazon
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
overall looks good, please see comments inline.
|
|
||
| Set the number of aggregation intervals from your data stream to consider in a detection window. We recommend you choose this value based on your actual data to see which one leads to the best results for your use case. | ||
|
|
||
| Based on experiments performed on a wide variety of one-dimensional data streams, we recommend using a window size between 1 and 16. The default window size is 8. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The default window size (shingle size at the backend?) is different for single-entity and high cardinality detector.
single-entity: 8
high cardinality: 4
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Sorry, seems we have changed to 1, but there is a TODO saying will change to 4, https://github.com/opendistro-for-elasticsearch/anomaly-detection/blob/5093c8a4df09fe665fe93eb6e5a994206adcfc2b/src/main/java/com/amazon/opendistroforelasticsearch/ad/settings/AnomalyDetectorSettings.java#L233
@kaituo can you confirm?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
yeah, high cardinality detector's shingle size is 1 now.
Issue #, if available:
Description of changes:
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.