🚀 New Feature Alert: Outlier Avoidance
Hey team!
Announcing the launch of our latest feature, "Outlier Avoidance," designed to enhance the accuracy and reliability of anomaly detection for Average type verses, such as AverageDrift, AverageOutlier, and AverageSuddenChange.
With this new feature, our anomaly detectors will intelligently avoid records where the value of a metric falls outside the specified percentile range, preventing outlier values from influencing the average calculations and generating unnecessary or skewed insights.
You can easily set the limits to determine outliers right from the metric chip itself in the verse GUI. By adjusting the bottom and top percentiles, you can define the desired range of values to check. Any values falling above or below this range will be avoided during analysis.
In the insights themselves, you'll find a special indication in the baseline segment chip, showing the range of values excluded from the analysis. This way, you have full visibility into the impact of "Outlier Avoidance" on your results.
We believe this addition will significantly improve the ease of use and save valuable time by avoiding false alarms.
As always, your feedback is essential to us. We'd love to hear your thoughts on this new feature. Should you have any questions or need assistance, our support team is here to help.