You are configuring a machine learning model for a client. They require a high degree of interpretability. What should you use?
AEnsemble methods
BDeep Neural Networks
CLinear Regression
DGradient Boosting
Explanation
Linear Regression is inherently interpretable, while others are more complex.
Q42
What happens when an ML model is trained with biased data?
AModel accuracy increases
BResults are more generalized
CPerformance is unchanged
DModel predictions may be unfair
Explanation
Biased data typically leads to unfair predictions, affecting model integrity.
Q43
Which Google Cloud service provides fully managed data warehousing?
ABigQuery
BCloud Storage
CCloud Spanner
DCloud SQL
Explanation
BigQuery is designed for data warehousing, while others serve different purposes.
Q44
A company needs to deploy a machine learning model with low latency for real-time predictions. Which Google Cloud service should they use?
ACloud ML Engine
BCloud Functions
CAI Platform Prediction
DDataflow
Explanation
AI Platform Prediction is specifically designed for real-time predictions, unlike the others, which have different uses.
Q45
You are configuring a Continuous Integration/Continuous Deployment (CI/CD) pipeline. What should you use to manage connections to Cloud services securely within the pipeline?
AService Account
BAPI Key
CEnvironment Variables
DIAM Policies
Explanation
Service Accounts manage secure access, while the others lack the appropriate security context for CI/CD.
Q46
Which service allows for seamless serverless machine learning model deployment?
AAI Platform Predictions
BCompute Engine
CCloud Functions
DApp Engine
Explanation
AI Platform Predictions is specifically designed for ML model deployment, while others serve different purposes.
Q47
A company needs to analyze large streams of data in real-time. Which Google Cloud service should they choose?
ABigQuery
BCloud Pub/Sub
CCloud Storage
DCloud Dataflow
Explanation
Cloud Pub/Sub is optimized for ingesting data in real-time, while others focus on batch processing or data storage.
Q48
What happens when you set a model's learning rate too high during training?
AIncreased accuracy until convergence
BOscillating loss during training
CFaster convergence guaranteed
DOverfitting of model occurs
Explanation
A high learning rate can cause oscillations in the loss, leading to poor model performance, while the other options suggest unrealistic outcomes.
Q49
Which service is best for scalable machine learning model deployment?
AAI Platform
BCloud Functions
CCloud Run
DGoogle Kubernetes Engine
Explanation
AI Platform is specifically designed for scalable machine learning deployments; the others serve different purposes.
Q50
A company needs to analyze streaming data in real-time. Which Google Cloud service should they use?
ABigQuery
BDataflow
CCloud Storage
DPub/Sub
Explanation
Dataflow is optimized for real-time stream processing, while the others are more suited for batch or static data analysis.