5 Classification Online Quiz Questions

A model uses message features to predict whether an email is “spam” or “not spam.” Which type of machine-learning task is it?

True or false: Despite its name, logistic regression is commonly used for classification.

Select all scenarios that illustrate multilabel classification.

Which classifier seeks a boundary that separates classes with a wide margin?

Which classifier predicts from the labels of nearby training examples, typically using a majority vote?

Select all statements that accurately describe the named classifiers.

True or false: If a multiclass model estimates probabilities for mutually exclusive classes, those estimates sum to 1.

Which theorem does Naive Bayes use to estimate class probabilities?

Which statement correctly distinguishes a classifier’s class prediction from its probability estimates?

A binary classifier estimates a spam probability of $0.82$ and a not-spam probability of $0.18$. Explain what these estimates do and do not establish, and distinguish them from a class prediction.

A team needs a classifier whose feature-based decision rules can be easy to inspect. Which method best fits that need?

An email system uses message words and sender details to predict whether each message is spam or not spam. Which kind of task is this?

True or false: A K-nearest neighbors classifier predicts a new example’s label using labels from nearby training examples, often by majority vote.

A photo-tagging system may assign both “beach” and “sunset” to the same photo. Which classification setup does this illustrate?

A team wants to combine predictions from many decision trees to often improve accuracy and help control overfitting. Which classifier should they consider?

A team needs a classifier whose sequence of feature-based rules can be inspected to understand decisions such as whether a message came from an unknown sender. Which classifier best fits this need?

A text classifier estimates class probabilities using Bayes’ theorem and assumes that features are conditionally independent given the class. What classifier is it?

A classification problem calls for a boundary with a wide margin, and the team may need a kernel to represent a more complex boundary. Which classifier is designed around this approach?