Skip to content

ML.NET Anomaly Detection Task output should include PredictedLabel #13481

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
colbylwilliams opened this issue Jul 19, 2019 · 3 comments · Fixed by #16388
Closed

ML.NET Anomaly Detection Task output should include PredictedLabel #13481

colbylwilliams opened this issue Jul 19, 2019 · 3 comments · Fixed by #16388

Comments

@colbylwilliams
Copy link
Member

The Anomaly detection inputs and outputs section should have PredictedLabel in addition to Score. Note: There is an issue with the current release of ML.NET where PredictedLabel always evaluates to true.

Additionally, the description for Score reads:

The non-negative, unbounded score that was calculated by the anomaly detection model.

This does not indicate how the value returned by the model for Score should be interpreted. Should higher scores be interpreted as anomalies, or scores closer to zero?


Document Details

Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

@colbylwilliams
Copy link
Member Author

Confirmed here, Predictions that are anomalies will have higher scores than those that aren't anomalies.

What it is essentially computing is 1 minus the ratio of the length of the projected vector and the input vector. We expect the higher the scores the higher the chance of an anomaly because in case of no anomaly the data point should be able to be projected without changing the length much.

@natke natke added triaged and removed ⌚ Not Triaged Not triaged labels Aug 23, 2019
@natke
Copy link
Contributor

natke commented Aug 23, 2019

Thanks @colbylwilliams. We will add this as part of the updates for #13880

@luisquintanilla
Copy link
Contributor

luisquintanilla commented Sep 10, 2019

To be worked on as part of AB#1596627

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging a pull request may close this issue.

5 participants