[ML] Improvements to sparse count modelling #721
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This simplifies sparse count and sum modelling and migrates to always updating the time series model, but using a weight which decreases in proportion to the number of empty buckets. This means we simply smoothly transition to modelling non-empty buckets for sparse data.
I've also removed the correction to the probability which accounts for the fraction of non-empty buckets: we have the rare function anyway if this is the primary concern. Finally, I changed periodicity testing so that it approximates the old behaviour, i.e. it tends to ignoring empty buckets as their proportion increases.
Closes #696.