You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
DetectEntireAnomalyBySrCnn - No anomalies detected as of version 1.5.4 or greater #5891
System Information (please complete the following information):
OS & Version: Windows 10
ML.NET Version: ML.NET 1.5.4 and greater
.NET Version: .NET 5
Describe the bug
When using DetectEntireAnomalyBySrCnn with versions 1.5.4 or greater, no anomalies are detected when running the DetectAnomaly method from the phone call tutorial.
Changing options does nothing. Has anything changed with this algorithm that would suddenly make it stop working after 1.5.4?
If you run the sample using version 1.5.2 with the same data and options it works.
changed the title [-]Changes to Anomaly Detection[/-][+]DetectEntireAnomalyBySrCnn - No anomalies detected as of version 1.5.4 or greater[/+]on Jul 28, 2021
Activity
[-]Changes to Anomaly Detection[/-][+]DetectEntireAnomalyBySrCnn - No anomalies detected as of version 1.5.4 or greater[/+]michaelgsharp commentedon Aug 19, 2021
@luisquintanilla since this is a regression I would say its probably a higher priority. @briacht thoughts?
briacht commentedon Aug 19, 2021
Yeah if the algorithm isn't working anymore we should definitely try and fix that
michaelgsharp commentedon Sep 27, 2021
This is the PR that caused this bug #5202.
matouskozak commentedon Sep 22, 2026
The current tests use sensitivity 87 nor 64 (
test/Microsoft.ML.TimeSeries.Tests/TimeSeriesDirectApi.cs)This sample program should be updated https://github.com/dotnet/samples/blob/main/machine-learning/tutorials/PhoneCallsAnomalyDetection/Program.cs
The code itself was fixed in #7610