Google Cloud

Google Cloud Certified – Professional Machine Learning Engineer

PR000269
Trending

Validate your skills as a Professional Machine Learning Engineer with exam code PR000269 in Google Cloud.

495 questions 0 views Free
Start Mock Test Timed · Full-length · Scored

Questions 261–270 of 495

Q261

What happens when a model is overfitting on training data?

  • A High accuracy on training data
  • B Low accuracy on validation data
  • C Balanced performance on both
  • D Increased training time required
Explanation Overfitting typically leads to high training accuracy but low validation accuracy due to memorization of noise in training data.
Q262

Which Google Cloud service is best for managing ML model lifecycle?

  • A AI Platform
  • B Cloud Functions
  • C Cloud Firestore
  • D BigQuery
Explanation AI Platform provides tools for model training, deployment, and versioning, while others serve different purposes.
Q263

A company needs real-time predictions for customer insights. Which should they use?

  • A Batch Processing with BigQuery
  • B AI Platform Prediction
  • C Cloud Pub/Sub
  • D Dataflow
Explanation AI Platform Prediction facilitates real-time inference, while others are focused on data flow or batch processes.
Q264

You are configuring a Neural Network in TensorFlow. What happens when the learning rate is set too high?

  • A It converges faster
  • B It may diverge
  • C It improves accuracy
  • D It trains slower
Explanation A high learning rate can cause the model to overshoot minima, leading to divergence rather than convergence.
Q265

Which service is best for deploying machine learning models at scale?

  • A AI Platform
  • B Cloud Functions
  • C BigQuery
  • D Cloud Dataflow
Explanation AI Platform is specifically designed for model deployment, whereas the others serve different purposes.
Q266

A company needs to process large batches of data daily and requires low latency responses. Which Google Cloud service should they use?

  • A Cloud Pub/Sub
  • B Cloud Storage
  • C BigQuery
  • D Dataflow
Explanation Dataflow can process batches with low latency, while the others are less optimized for this specific use case.
Q267

What happens when you use a pre-trained TensorFlow model that exceeds the memory limit of your VM?

  • A Model fails to load
  • B Auto-scaling adjusts resources
  • C Model runs slower automatically
  • D It's cached in memory
Explanation The model cannot load if it exceeds the VM's memory capacity; the other options do not accurately describe the behavior in this scenario.
Q268

Which Google Cloud service helps in real-time data streaming?

  • A Cloud Functions
  • B Cloud Pub/Sub
  • C Cloud Run
  • D BigQuery
Explanation Cloud Pub/Sub is designed for real-time messaging, while the others serve different purposes.
Q269

A company needs to run machine learning models on large datasets while keeping costs low. Which service should they choose?

  • A AI Platform Notebooks
  • B BigQuery ML
  • C TensorFlow Serving
  • D Cloud Dataflow
Explanation BigQuery ML allows efficient ML on large datasets directly in BigQuery, unlike the other options that are more resource-intensive.
Q270

You are configuring a machine learning model for long-term predictions. What happens when you don't account for seasonality in your data preprocessing?

  • A Reduced model complexity
  • B Improved accuracy
  • C Underfitting of predictions
  • D Overfitting of predictions
Explanation Ignoring seasonality may lead to underfitting, as the model won't capture periodic patterns.