PMLE Practice Exam 60 — 20 Questions

Question 1

A healthcare organization is building a machine learning model to predict patient readmission risk. They have sensitive data stored in BigQuery that includes protected health information (PHI). The data science team uses Vertex AI Workbench notebooks to explore the data and develop models. The organization's security policy requires that all PHI data must be encrypted at rest and in transit, and that access to the data is logged and audited. They also need to ensure that the data used for model training is de-identified to remove direct identifiers such as patient names and SSNs. The team wants to automate the de-identification process as part of the data pipeline. Which approach meets these requirements?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Create a Dataflow pipeline that reads from the original BigQuery table, applies Cloud DLP de-identification transforms, and writes to a new BigQuery table. Grant the data science team access to the de-identified table.. It uses Cloud DLP within a Dataflow pipeline to automatically de-identify PHI data as it is read from the original BigQuery table and written to a new, de-identified table. This satisfies the requirement for automated de-identification, while the original table remains encrypted at rest (BigQuery default) and in transit (TLS), and access to the original data can be logged via Cloud Audit Logs. The data science team only gets access to the de-identified table, ensuring PHI is not exposed during model development.

Answer

Enable Shielded VM on Vertex AI Workbench notebooks and use VPC-SC to restrict data access.

Answer

Use Cloud Key Management Service to encrypt the PHI columns in BigQuery, and share the encryption key with the data science team.

Answer

Use BigQuery row-level security to mask PHI columns for the data science team, and train the model directly on the original table.

Question 2

A data science team is configuring Vertex AI Model Monitoring for a deployed model. They want to detect both feature skew and feature drift. Which TWO configurations must they set?

Accepted Answer

Configure the monitoring frequency (e.g., hourly).. To detect feature skew, they need to specify a training dataset or statistics as a reference. To detect drift over time, they need to set a monitoring frequency. Sampling rate is optional but recommended.

Answer

Enable request/response logging.

Answer

Select the Jensen-Shannon divergence algorithm.

Answer

Set the sampling rate to 1.0 (100%).

Question 3

Refer to the exhibit. An engineer notices no drift alerts but the model performance has degraded. What is the likely cause?

Accepted Answer

Concept drift is occurring, which is not captured by drift or skew detection. Concept drift occurs when the statistical properties of the target variable change over time, causing model performance to degrade even when the input data distribution remains stable. Drift detection (e.g., data drift or skew) monitors changes in feature distributions, not the relationship between features and the target. Since no drift alerts were triggered, the input data appears unchanged, but the model's predictive relationship has shifted — this is classic concept drift, which requires performance monitoring (e.g., accuracy, F1-score) rather than drift or skew detection.

Answer

Feature attribution monitoring is causing too many false positives

Answer

Drift threshold for income is too high

Answer

Skew thresholds are not configured for categorical features

Question 4

Refer to the exhibit. The team notices that the pipeline fails to read data from the specified Cloud Storage path. What is the most likely issue?

Accepted Answer

The service account lacks `storage.objectViewer` permission. The pipeline fails to read data from Cloud Storage because the service account lacks the `storage.objectViewer` IAM role, which grants the `storage.objects.get` and `storage.objects.list` permissions required to read objects. Without this role, the pipeline cannot authenticate or authorize the read operation, even if the bucket and path are correct.

Answer

The bucket does not exist

Answer

The pipeline runner is incorrect

Answer

The region is mismatched

Question 5

You manage a multi-tenant serving system on Vertex AI Prediction where multiple models are deployed in a single endpoint using model versioning. One particular model version (v2) is consuming excessive resources, causing latency spikes for other versions. You need to isolate this model to prevent interference. The models are all in TensorFlow SavedModel format. What is the best approach?

Accepted Answer

Use Vertex AI Model Registry to deploy v2 to a dedicated endpoint and update the model alias.. The best approach is Option C: Use Vertex AI Model Registry to deploy v2 to a dedicated endpoint and update the model alias. This isolates v2's resource consumption without affecting other versions. Option A involves custom routing logic which adds complexity and does not guarantee isolation. Option B is incorrect because Vertex AI Prediction does not support setting resource limits on containers directly. Option D is less ideal as it requires managing a separate load balancer and manual traffic splitting, whereas Vertex AI Model Registry allows seamless alias-based routing.

Answer

Shard the models across multiple replicas using a custom routing logic in the container.

Answer

Set resource limits on the container using Kubernetes resource requests/limits, but Vertex AI Prediction does not support that.

Answer

Create a separate endpoint for v2 and redirect traffic using a load balancer.

Question 6

You are deploying a new version of a model to a Vertex AI endpoint that already has a champion model serving 100% of traffic. You want to gradually shift traffic to the new version while monitoring for errors. Which approach should you use?

Accepted Answer

Deploy the challenger to the same endpoint with initial traffic split, e.g., champion 90%, challenger 10%, and gradually adjust.. Vertex AI endpoints support traffic splitting between model versions deployed to the same endpoint. By deploying the challenger to the same endpoint and setting an initial split (e.g., champion 90%, challenger 10%), you can gradually shift traffic while monitoring for errors. This approach uses the endpoint's built-in traffic management, avoiding the complexity and latency of external load balancers.

Answer

Use Cloud Load Balancing with weighted backend services pointing to different endpoints.

Answer

Delete the champion model and redeploy with the challenger as the new version.

Answer

Create a new endpoint for the challenger and use a load balancer to split traffic.

Question 7

A company wants to classify support ticket text into categories. They have labeled historical tickets. Which Google Cloud service allows them to train a custom classification model with no code?

Accepted Answer

AutoML Natural Language. AutoML Natural Language (now part of Vertex AI) is the correct service because it enables users to train custom text classification models using labeled data without writing any code. It provides a no-code interface for uploading datasets, training models, and evaluating performance, making it ideal for classifying support ticket text into custom categories.

Answer

Vertex AI Matching Engine

Answer

Cloud Natural Language API

Answer

Document AI

Question 8

A data science team wants to build a Vertex AI pipeline that trains a model, evaluates it, and conditionally deploys it if the accuracy exceeds 0.9. They want to use the Kubeflow Pipelines SDK v2. Which construct allows them to conditionally execute the deployment step based on the evaluation metric?

Accepted Answer

dsl.If. In Kubeflow Pipelines SDK v2, `dsl.If` is the correct construct for conditionally executing pipeline steps based on runtime metrics or parameters. It allows you to define a condition that, when evaluated to true, triggers the deployment step only if the model accuracy exceeds 0.9. This is the standard way to implement branching logic in v2 pipelines.

Answer

dsl.Conditional

Answer

dsl.ExitHandler

Answer

dsl.Collected

Question 9

Which TWO of the following can be used as input sources for Vertex AI batch prediction jobs? (Choose 2)

Accepted Answer

BigQuery. BigQuery is a supported input source for Vertex AI batch prediction jobs because Vertex AI can directly read data from BigQuery tables for batch predictions. This integration allows you to store your prediction requests in BigQuery and have Vertex AI write the predictions back to a BigQuery output table, streamlining the workflow for large-scale predictions without needing to export data to Cloud Storage first.

Answer

Cloud Firestore

Answer

Cloud SQL

Answer

Cloud Spanner

Question 10

A financial services company uses a custom deep learning model on Vertex AI to automatically approve or reject credit card transactions. The model is explainable using Vertex Explainable AI, and the company monitors feature attribution drift with thresholds defined per feature. Last week, the monitoring system flagged that the mean absolute attribution score for the 'transaction_amount' feature increased from 0.35 to 0.55. The overall model accuracy, measured on a daily batch of labeled transactions, has remained around 97%. The operations team is concerned about potential compliance issues due to changing model behavior. What should the data scientist do?

Accepted Answer

Investigate whether there has been a shift in the distribution of 'transaction_amount' values in the recent transaction data, which could explain the attribution change.. A shift in the distribution of the 'transaction_amount' feature (e.g., due to seasonality or a new customer segment) can naturally cause its attribution score to change without indicating model degradation. Vertex Explainable AI computes feature attributions relative to the current data distribution; if the input values shift, the model's reliance on that feature may legitimately increase. Investigating the distribution shift is the first diagnostic step before adjusting thresholds or retraining, as stable accuracy does not rule out data drift that could lead to compliance issues.

Answer

Tune the alert threshold for 'transaction_amount' to 0.6 to avoid future false alarms.

Answer

Retrain the model by increasing regularization to reduce the importance of the 'transaction_amount' feature.

Answer

Disable the feature attribution drift monitoring for 'transaction_amount' since the model accuracy is stable.

Question 11

A company wants to use low-code ML for time series forecasting with 5 years of hourly data. They need to incorporate holiday effects. Which solution best meets these requirements?

Accepted Answer

BigQuery ML ARIMA_PLUS with holiday regression. BigQuery ML ARIMA_PLUS with holiday regression is the correct choice because it is a low-code solution that natively supports time series forecasting with built-in holiday effect modeling. ARIMA_PLUS automatically handles seasonality, trend, and holiday regression without requiring custom code, making it ideal for 5 years of hourly data.

Answer

Custom LSTM model

Answer

Vertex AI AutoML Tables with timestamp and holiday features

Answer

Vertex AI AutoML Forecasting with timestamp and holiday feature

Question 12

An engineer wants to use BigQuery ML to explain predictions from a trained boosted tree classifier for a specific set of input rows. Which function should they use?

Accepted Answer

ML.EXPLAIN_PREDICT. ML.EXPLAIN_PREDICT provides feature attributions for each prediction. ML.EVALUATE returns aggregate metrics, ML.FEATURE_IMPORTANCE gives global importance, and ML.PREDICT only returns predictions.

Answer

ML.EVALUATE

Answer

ML.FEATURE_IMPORTANCE

Answer

ML.PREDICT

Question 13

Refer to the exhibit. A team runs this command to upload a model to Vertex AI. They want to create this model as a new version under an existing model named 'my_model'. What is missing from the command?

Accepted Answer

--parent-model=my_model. The `--parent-model` flag is required when uploading a new model version to an existing model in Vertex AI. Without specifying the parent model name, the command would attempt to create a brand-new model rather than adding a version to the existing 'my_model'. The `gcloud ai models upload` command uses this flag to associate the new version with the specified parent model.

Answer

--description='Second version'

Answer

--version=v2

Answer

--labels=team=ml

Answer

--service-account=sa@project.iam.gserviceaccount.com

Question 14

A startup is deploying its first machine learning model using BigQuery ML. The model is a logistic regression for churn prediction, trained on a dataset of 5 million rows. The pipeline runs every week: it exports training data from BigQuery, trains a model using BigQuery ML, and then deploys the model as a remote model for predictions. The ML engineer wants to set up basic monitoring to ensure the pipeline runs successfully and the model quality does not degrade. Which monitoring approach should the engineer implement first?

Accepted Answer

Set up Cloud Monitoring alerts on the pipeline's execution status and duration, and create a simple dashboard showing these metrics.. The first priority in monitoring a new ML pipeline is ensuring it runs successfully and on time. Cloud Monitoring alerts on execution status and duration directly address pipeline reliability, which is the most basic operational concern before model quality metrics like AUC or drift can be meaningful. This approach aligns with the principle of starting with infrastructure health before advanced model monitoring.

Answer

Export BigQuery audit logs to Cloud Logging and analyze them for any errors.

Answer

Enable Vertex AI Model Monitoring to detect data drift between training and serving data.

Answer

Monitor the model's area under the ROC curve (AUC) over time and alert if it drops by more than 0.01.

Question 15

A healthcare startup is developing a diagnostic model using sensitive patient data. They use Vertex AI to manage the training pipeline. They need to ensure that the data is encrypted both at rest and in transit. Additionally, they want to prevent the ML engineers from seeing raw data but still allow them to train models. They use Cloud Storage with CMEK and VPC-SC. They plan to use Vertex AI Training with a custom service account. The data stored in Cloud Storage is encrypted with CMEK. What additional step is needed to allow Vertex AI Training to access the encrypted data?

Accepted Answer

Grant the custom service account the Cloud KMS CryptoKey Decrypter role.. Vertex AI Training must use a custom service account that has the Cloud KMS CryptoKey Decrypter role to decrypt the CMEK-encrypted data stored in Cloud Storage. The custom service account is the identity that Vertex AI jobs run as, and it needs explicit permission to decrypt the CMEK key to read the training data. Without this role, the encrypted objects remain inaccessible even if the service account has Storage Object Viewer permissions.

Answer

Use a service account with the 'Storage Admin' role and 'Cloud KMS CryptoKey Decrypter' role.

Answer

Grant the Cloud Storage service agent the Cloud KMS CryptoKey Decrypter role.

Answer

Disable encryption for the training data to simplify access.

Question 16

You are defining a Python function component in KFP SDK v2. Which decorator should you use?

Accepted Answer

@dsl.component. In KFP SDK v2, the `@dsl.component` decorator is used to define a Python function as a lightweight, reusable pipeline component that can be executed independently. This decorator automatically generates a containerized component from the function's signature and type annotations, enabling type-safe inputs and outputs without requiring a separate component YAML specification.

Answer

@dsl.task

Answer

@component

Answer

@dsl.pipeline

Question 17

You are optimizing a model for deployment on Vertex AI using NVIDIA Triton Inference Server. Which TWO actions can you take to improve inference performance?

Accepted Answer

Use TensorRT to quantize the model to FP16 or INT8.. TensorRT optimizes model inference by quantizing weights and activations to lower precision formats like FP16 or INT8, reducing memory bandwidth and computation time without significant accuracy loss. This is a standard technique for improving throughput on NVIDIA GPUs, especially when deploying with Triton Inference Server, which natively supports TensorRT-optimized model repositories.

Answer

Increase the number of model replicas to the maximum.

Answer

Disable model caching to reduce memory usage.

Answer

Use a larger machine type with more vCPUs.

Question 18

Drag and drop the steps to set up model monitoring for drift detection on Vertex AI in the correct order.

Accepted Answer

Deploy model, then Enable monitoring, then Set thresholds, then Configure alerts, then Review. The correct order to set up model monitoring for drift detection on Vertex AI is: 1. Deploy model to an endpoint. 2. Enable monitoring on the deployed endpoint. 3. Set drift detection thresholds (e.g., for feature distributions or prediction outputs). 4. Configure alerts (e.g., email notifications). 5. Review monitoring results and alerts.

Answer

Deploy model, then Set thresholds, then Enable monitoring, then Configure alerts, then Review

Answer

Enable monitoring, then Deploy model, then Set thresholds, then Configure alerts, then Review

Answer

Deploy model, then Enable monitoring, then Configure alerts, then Set thresholds, then Review

Question 19

Refer to the exhibit. A team deploys a model with the above configuration. They observe that during traffic spikes, the endpoint does not scale up quickly enough, causing increased latency. The average CPU utilization never exceeds 50%. What is the most likely reason for the slow scaling?

Accepted Answer

The autoscaling metric is not configured. The exhibit shows that autoscaling is not configured (or the autoscaling metric is missing). Without a configured autoscaling metric, the endpoint cannot automatically add instances in response to traffic spikes, causing slow scaling and increased latency. Option B is incorrect because if minReplicaCount were too low, it would limit the minimum number of replicas but would not prevent scaling up if autoscaling were active. Option C is incorrect because an accelerator bottleneck would manifest as high utilization or errors, not slow scaling with low CPU. Option D is incorrect because the average CPU utilization is below 50%, indicating the machine type has sufficient CPU capacity.

Answer

The minReplicaCount is too low

Answer

The accelerator is causing a bottleneck

Answer

The machineType does not have enough CPU

Question 20

A machine learning team has a prototype using a custom TensorFlow model trained on a small dataset stored in Cloud Storage. They want to scale the prototype to production with minimal code changes while ensuring the model can handle increased traffic and new data. The model currently loads data using tf.data.Dataset from CSV files. Which approach best meets these requirements?

Accepted Answer

Use Vertex AI Training with hyperparameter tuning and distributed training, then deploy the model to Vertex AI Prediction with autoscaling.. Vertex AI Prediction with autoscaling directly addresses the need to handle increased traffic without code changes, while Vertex AI Training with hyperparameter tuning and distributed training enables scaling to larger datasets with minimal modifications to the existing tf.data pipeline. This approach keeps the custom TensorFlow model intact and leverages managed infrastructure for both training and serving.

Answer

Deploy the model to AI Platform (Unified) Prediction with a custom container, and use AI Platform Training to retrain on larger datasets.

Answer

Migrate the model to BigQuery ML and use SQL for training and prediction to leverage BigQuery's scalability.

Answer

Package the model as a Cloud Run Function and use Cloud Scheduler to trigger retraining periodically.