Project 2 Midterm Prep Quiz

What does convergence mean in the context of gradient descent?

• When the loss function becomes exactly zero

• When the learning rate becomes constant

• When the optimization process stops improving significantly

• When weights increase indefinitely

What is a global minimum in an optimization problem?

• The lowest value in a local region of the loss function

• The average of all local minimums

• The highest point on the cost curve

• The absolute lowest value of the loss function across all parameter space

What is a local minimum?

• A temporary value reached before the global minimum

• A point where the gradient is zero, but not the absolute minimum

• A saddle point

• The lowest value that gradient descent can never reach

6. Which statement best describes L1 regularization (Lasso)?

• It adds the absolute value of coefficients as a penalty term to the loss function.

• It adds the square of coefficients as a penalty term to the loss function.

• It reduces the learning rate during training.

• It increases the number of features in the model.

What is the purpose of regularization in machine learning?

• To reduce training time

• To increase model complexity

• To reduce overfitting

• To eliminate outliers from the dataset

Which regularization method uses the square of the coefficients?

• Lasso

• Ridge

• Elastic Net

• None of the above

What is Elastic Net regularization?

• A method that switches between L1 and L2 during training

• A combination of L1 and L2 regularization

• A regularization that focuses on maximizing entropy

• A method used to increase recall

What does the decision boundary or threshold represent in classification?

• The highest accuracy achieved

• The dividing line between different predicted classes

• The level of regularization applied

• The convergence point of gradient descent

What is the dividing line between different predicted classes?

• The threshold value

• The dividing line between different predicted classes

• The confusion matrix

• The ROC curve

What is the formula for accuracy in classification metrics?

• (TP + FN) / (TP + FP + TN + FN)

• (TP + TN) / (TP + TN + FP + FN)

• TP / (TP + FP)

• TP / (TP + FN)