5 Classification Free Online FlashCards

What does classification predict?

Classification is a prediction task in which a model assigns an example to one of a set of categories.

What defines binary classification?

Binary classification has two possible classes, such as spam and not spam.

What are features in a classification task?

Features are properties of an example that the model uses to make a prediction; in email classification, words can be features.

How does classification differ from regression?

Classification predicts a category; regression predicts a numerical quantity.

What is a classifier's decision boundary?

A decision boundary is where a classifier's predicted class changes, separating feature-space regions assigned to different classes.

How does a decision tree classify an example?

A decision tree applies a sequence of feature-based rules or splits to reach a class prediction.

How does logistic regression turn a probability into a binary label?

Logistic regression estimates class probabilities. For a binary task, a chosen threshold converts a probability into one of the two labels.

In classification, what is a class?

A class is one of the categories a classifier can assign to an example.

What defines multiclass classification?

Multiclass classification has more than two possible classes, such as cat, dog, and bird.

What role do labeled training examples play?

The model learns from training examples that already have labels, then uses what it learns to classify new examples.

Why might a decision boundary be hard to draw?

With many features, a decision boundary lies in a higher-dimensional feature space and may not be drawable on a page.

What kind of boundary can a linear classifier make in a two-feature plot?

A linear model or linear SVM can separate classes with a straight boundary in a plot with two features.