4 ms·
Ah but then you just move the error case to outlier detection.
by moffkalast 7mo ago
Ah but then you just move the error case to outlier detection.
- alex_be 7mo agoTrue. It's about managing the risk rather than eliminating it. If you remove an outlier, you get a missing measurement and, as a result, higher uncertainty (error). But it is still better than keeping the outlier.