Question 1

Which approach is used in object detection to ensure that small objects are detected accurately in an image with a high resolution?

Options :

A : Single-Scale Detection

B : Multi-Scale Detection

C : Image Rescaling

D : Class Balancing

Answer: B

Question 2

In which scenario would you use Batch Normalization?

Options :

A : To perform feature selection by reducing dimensionality.

B : To balance the training and test dataset.

C : To stabilize and speed up training by normalizing activations.

D : To increase the number of model parameters.

Answer: C

Question 3

How can ensemble methods improve the robustness of large language models in production environments?

Options :

A : By refining the tokenization process

B : By optimizing the learning rate schedules

C : By increasing the training data size

D : By combining outputs of multiple models

Answer: D

Question 4

Why is the sigmoid function crucial in binary logistic regression?

Options :

A : It smooths the decision boundary for non-linear data.

B : It provides feature scaling to normalize input features.

C : It maps the output to probabilities between zero and one.

D : It adjusts the learning rate for better convergence speed.

Answer: C

Question 5

Which strategy is used to mitigate the risk of biased outputs in large language models?

Options :

A : Regularly updating the model-s weights

B : Conducting adversarial training

C : Increasing model interpretability

D : Implementing gradient clipping

Answer: B

Higher Test Marks with Free Online CAIS Exam Practice