Data Science Supervised Learning Algorithms 3

Question 1

In random forests, what technique reduces correlation between individual trees?

Accepted Answer

Random feature subsampling at each split

Answer

Pruning every tree

Answer

Using a single deep tree

Answer

Increasing the learning rate

Question 2

What is the key difference between bagging and boosting?

Accepted Answer

Bagging trains models independently in parallel; boosting trains sequentially focusing on errors

Answer

Bagging trains models sequentially; boosting trains in parallel

Answer

Both train identically

Answer

Boosting only uses one model

Question 3

In gradient boosting, what do successive trees primarily fit?

Accepted Answer

The residual errors of previous trees

Answer

The raw target values

Answer

Random noise

Answer

The feature means

Question 4

What does a large C value in an SVM control?

Accepted Answer

A narrower margin penalizing misclassifications more

Answer

A wider margin with more misclassifications allowed

Answer

The kernel type

Answer

The number of support vectors only

Question 5

Which loss function is typically minimized in standard linear regression?

Accepted Answer

Mean squared error

Answer

Cross-entropy loss

Answer

Hinge loss

Answer

Log loss

Question 6

What problem does L1 (Lasso) regularization help address that L2 does not?

Accepted Answer

It can drive some coefficients exactly to zero, performing feature selection

Answer

It always increases bias

Answer

It prevents any overfitting

Answer

It speeds up training only

Question 7

In a confusion matrix, what does recall measure?

Accepted Answer

True positives over all actual positives

Answer

True positives over all predicted positives

Answer

True negatives over all negatives

Answer

Overall correct predictions