GMLE by GIAC Actual Free Exam Q&As

Question 1

Which of the following are important when using standard deviation to understand data spread?
(Choose two)
Response:

A. Standard deviation is not influenced by outliers B. Standard deviation can only be used for categorical data C. A large standard deviation means that data points are more spread out D. A small standard deviation means that data points are close to the mean

Discussion 0

Correct Answer: C,D Vote an answer

Question 2

What is the main advantage of using a 'bagging' approach in ensemble learning?
Response:

A. To enhance the speed of training models B. To reduce the variance of the model C. To optimize individual model parameters D. To increase the bias in the model

Discussion 0

Correct Answer: B Vote an answer

Question 3

Which of the following are key characteristics of genetic algorithms used in neural network optimization?
(Choose two)
Response:

A. Random initialization of weights B. Evolution through mutation and crossover C. Manual selection of hyperparameters D. Fitness evaluation based on a predefined objective

Discussion 0

Correct Answer: B,D Vote an answer

Question 4

Which strategies can be applied to improve the anomaly detection performance of an autoencoder?
(Choose two)
Response:

A. Add regularization techniques such as L1 or L2 B. Increase the size of the latent space C. Train the autoencoder on both normal and anomalous data D. Reduce the number of hidden layers

Discussion 0

Correct Answer: A,C Vote an answer

Question 5

What does 'grid search' refer to in hyperparameter tuning?
Response:

A. A method to visualize the performance of different models B. The process of randomly selecting hyperparameters for testing C. A strategy for reducing the number of hyperparameters D. A technique to search exhaustively through a specified subset of hyperparameters

Discussion 0

Correct Answer: D Vote an answer

Question 6

Stochastic Gradient Descent differs from traditional Gradient Descent by:
Response:

A. Updating model parameters after evaluating each data point B. Eliminating the need for a learning rate C. Using a fixed learning rate throughout the training process D. Updating model parameters after evaluating the entire dataset

Discussion 0

Correct Answer: A Vote an answer

Question 7

In machine learning, what is 'model validation'?
Response:

A. The technique of testing the model on unseen data B. The practice of combining multiple models C. The process of training the model D. The method of improving the model's performance

Discussion 0

Correct Answer: A Vote an answer

Question 8

In machine learning, what is 'data normalization'?
Response:

A. Dividing the dataset into training and test sets B. Removing duplicate records from the dataset C. Scaling numeric features to a standard range D. Encoding categorical variables into numerical values

Discussion 0

Correct Answer: C Vote an answer

Question 9

Which of the following is commonly used to prevent overfitting in deep neural networks?
Response:

A. Decreasing the dataset size B. Increasing the learning rate C. Using dropout during training D. Increasing the number of hidden layers

Discussion 0

Correct Answer: C Vote an answer

GIAC Machine Learning Engineer - GMLE Exam Practice Test