Changing the Tusk on a Noisy Test

Going from the threshold base system to the change point detection dropped our false positives by at least an order of magnitude, if not more. It was drastically better. But we still have more not tura ven in call false positives. We still flag more things than want to look at. And so we're continually trying to improve that emake sense. If any of your listeners have ideas on all ears, but we have cases of orrelated noise.