nerdexam
Google

PROFESSIONAL-MACHINE-LEARNING-ENGINEER · Question #239

You are investigating the root cause of a misclassification error made by one of your models. You used Vertex AI Pipelines to train and deploy the model. The pipeline reads data from BigQuery…

The correct answer is B. Use the lineage feature of Vertex AI Metadata to find the model artifact. Determine the version of. To recover the training data for a misclassifying model version deployed via Vertex AI Pipelines, leverage Vertex AI Metadata's lineage feature.

Submitted by rachelw· Apr 18, 2026ML pipeline operationalization

Question

You are investigating the root cause of a misclassification error made by one of your models. You used Vertex AI Pipelines to train and deploy the model. The pipeline reads data from BigQuery. creates a copy of the data in Cloud Storage in TFRecord format, trains the model in Vertex AI Training on that copy, and deploys the model to a Vertex AI endpoint. You have identified the specific version of that model that misclassified, and you need to recover the data this model was trained on. How should you find that copy of the data?

Options

  • AUse Vertex AI Feature Store. Modify the pipeline to use the feature store, and ensure that all
  • BUse the lineage feature of Vertex AI Metadata to find the model artifact. Determine the version of
  • CUse the logging features in the Vertex AI endpoint to determine the timestamp of the model's
  • DFind the job ID in Vertex AI Training corresponding to the training for the model. Search in the

How the community answered

(54 responses)
  • A
    4% (2)
  • B
    83% (45)
  • C
    2% (1)
  • D
    11% (6)

Why each option

To recover the training data for a misclassifying model version deployed via Vertex AI Pipelines, leverage Vertex AI Metadata's lineage feature.

AUse Vertex AI Feature Store. Modify the pipeline to use the feature store, and ensure that all

Vertex AI Feature Store manages and serves features for model training and serving, but it does not provide lineage tracking to recover historical training data for an already trained and deployed model.

BUse the lineage feature of Vertex AI Metadata to find the model artifact. Determine the version ofCorrect

Vertex AI Metadata provides lineage tracking that records the relationships between datasets, training jobs, and model artifacts. By locating the specific misclassifying model version's artifact, its lineage can be traced back to identify the exact copy of the training data in Cloud Storage used to train it.

CUse the logging features in the Vertex AI endpoint to determine the timestamp of the model's

Logging features in a Vertex AI endpoint primarily capture inference requests and responses, not the historical training data associated with the deployed model version.

DFind the job ID in Vertex AI Training corresponding to the training for the model. Search in the

While finding the job ID is a step, directly searching Cloud Storage by job ID lacks the structured, queryable linkage that Vertex AI Metadata provides for definitively associating a model with its specific training data artifact.

Concept tested: Vertex AI Metadata lineage tracking

Source: https://cloud.google.com/vertex-ai/docs/ml-metadata/overview#lineage_tracking

Topics

#Vertex AI Metadata#ML Lineage#ML Pipeline Operationalization#Model Debugging

Community Discussion

No community discussion yet for this question.

Sign up to join the discussion
Full PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice