Data Science Supervised Learning Models 4

Question 1

What does a ROC curve plot?

Accepted Answer

True positive rate vs false positive rate

Answer

Precision vs recall

Answer

Accuracy vs epochs

Answer

Loss vs learning rate

Question 2

An AUC of 0.5 indicates what about a classifier?

Accepted Answer

Performance no better than random guessing

Answer

Perfect classification

Answer

Severe overfitting

Answer

Zero variance

Question 3

What is cross-validation primarily used for?

Accepted Answer

Estimating model performance on unseen data

Answer

Increasing dataset size

Answer

Removing outliers

Answer

Encoding categorical variables

Question 4

Which technique handles categorical features for a linear model?

Accepted Answer

One-hot encoding

Answer

Min-max scaling

Answer

PCA

Answer

Dropout

Question 5

What does early stopping prevent during training?

Accepted Answer

Overfitting by halting when validation loss stops improving

Answer

Underfitting from too few epochs

Answer

Feature leakage

Answer

Class imbalance

Question 6

In random forests, how is diversity among trees achieved?

Accepted Answer

Bootstrap sampling and random feature subsets

Answer

Using a single deep tree

Answer

Sequential error correction

Answer

Shared training data only

Question 7

What is data leakage in supervised learning?

Accepted Answer

Information from outside the training set influencing the model

Answer

Loss of training data

Answer

A network security flaw

Answer

Missing values in features