There have been a couple of occasions where I've needed to identify when something is out of whack, for example checking an incoming set of market movement data for suspect quality issues so it was interesting to see the post on the Twitter Engineering blog about their open source time series anomaly detection package for R.
Be interesting to see how real time this could be. Given that it handles seasonality it could be a great tool to use in monitoring a checkout funnel.
Comments
You can follow this conversation by subscribing to the comment feed for this post.