Sagemaker Model Monitoring - Model Quality for Isolation forest model

Question

I am having an unsupervised learning model - isolation forest model which gives me two value 1, -1 (anomaly detection).

I want to build model quality on this. However I see it only support BinaryClassification, Multiclass and Regrression.

Can I still use Binaryclassification for this or MulticlassClassfication for building model quality on it?

Answer 1

You can do both Binary and MultiClass classification with Model Quality job - Please refer to the Model Quality Metrics support by SageMaker here https://docs.aws.amazon.com/sagemaker/latest/dg/model-monitor-model-quality-metrics.html