Question 1

What is an effective way to communicate risk status to stakeholders?

Options :

A : Providing detailed technical risk assessments

B : Only reporting risks that have already materialized

C : Using clear and concise summaries of risk status

D : Avoiding risk communication to prevent concern

Answer: C

Question 2

Which challenge is associated with deploying large language models on edge devices?

Options :

A : The computational resources required exceed those available on edge devices

B : The model-s inference speed decreases significantly

C : The model-s architecture needs to be simplified, reducing its effectiveness

D : The model-s ability to perform zero-shot tasks is lost

Answer: A

Question 3

Which evaluation metric is most suitable for imbalanced datasets?

Options :

A : F1 Score

B : Precision

C : Mean Squared Error

D : Accuracy

Answer: A

Question 4

What is the primary objective of a Denoising Autoencoder?

Options :

A : To reconstruct clean data from noisy inputs

B : To add noise to the input for robust learning

C : To compress data

D : To handle missing values

Answer: A

Question 5

What is the key advantage of using the RMSprop optimizer over standard SGD?

Options :

A : It reduces the model-s training time by half.

B : It eliminates the need for tuning the learning rate.

C : It adapts learning rates for each parameter based on recent gradient magnitudes.

D : It provides exact solutions to optimization problems.

Answer: C