Which service is best for training machine learning models using TensorFlow?
AGoogle AI Platform
BGoogle Compute Engine
CGoogle App Engine
DGoogle Cloud Functions
Explanation
Google AI Platform is specifically designed for ML model training, while others are general compute services.
Q32
A company needs to deploy a machine learning model for real-time prediction. Which service should they use?
AGoogle Cloud Run
BGoogle Cloud Storage
CGoogle Cloud Pub/Sub
DGoogle BigQuery
Explanation
Google Cloud Run allows for fast deployment of models for real-time prediction.
Q33
What happens when you set a high learning rate for a model?
AModel converges quickly
BModel training is slower
CModel may not converge
DModel error decreases
Explanation
A high learning rate can cause the model weights to oscillate or diverge, preventing convergence.
Q34
Which service is best for building event-driven systems?
ACloud Functions
BCompute Engine
CCloud Storage
DBigQuery
Explanation
Cloud Functions is designed for event-driven applications, while the others serve different purposes.
Q35
A company needs to analyze streaming data with low latency. Which service should they use?
ACloud Pub/Sub
BCloud Storage
CBigQuery
DCloud Spanner
Explanation
Cloud Pub/Sub is optimized for real-time message ingestion and delivery, unlike the other options suitable for batch or static data.
Q36
What happens when using an AI model trained on biased data?
AAccuracy improves significantly
BPerformance fails on all data
CIt displays biased predictions
DNo impact on predictions
Explanation
Models trained on biased data tend to replicate those biases in their predictions, affecting fairness and accuracy.
Q37
Which Google Cloud service offers serverless data processing?
AGoogle Cloud Functions
BGoogle Cloud Run
CGoogle Dataflow
DGoogle Compute Engine
Explanation
Google Dataflow specializes in serverless data processing; Functions and Run are for different use cases, while Compute Engine requires provisioned infrastructure.
Q38
A company needs to perform image classification using machine learning. Which API would they most likely use?
AVision AI
BNatural Language API
CSpeech-to-Text API
DTranslation API
Explanation
Vision AI is designed specifically for image classification, while the others serve different functionalities.
Q39
You are configuring a machine learning model in Vertex AI. What happens when you set the 'max parallelism' to a value greater than the dataset size?
AGPU resource allocation improves
BIt will run with that max value
CConcurrency is ignored
DPerformance is halved
Explanation
Setting 'max parallelism' above dataset size means concurrency cannot exceed the number of data items available, therefore it's ignored.
Q40
Which service is best for real-time data processing?
ACloud Dataflow
BCloud Storage
CBigQuery
DCloud Pub/Sub
Explanation
Cloud Dataflow is optimized for real-time stream processing, unlike the others.