A Classification Tree is a tree that predicts a categorical outcome.

A Classification Tree is a tree that predicts a categorical outcome.

• A. Predicts a continuous value
• B. Predicts a category
• C. Predicts a probability
• D. Predicts a time-to-event

Answer: (B)

Classification trees are built to assign observations to categories. They split the data in ways that group similar cases together, and each leaf of the tree ends with a label from a predefined set of categories. So, given a new observation, you follow the branches based on its feature values and land on a leaf that provides the predicted category. This is what makes them ideal for predicting categorical outcomes. If the goal were to predict a numeric value, a regression tree would be used instead, predicting a continuous quantity. A tree can also provide class probabilities by looking at the proportion of training examples of each category in a leaf, but the final prediction for an observation is typically the category with the highest probability. Time-to-event outcomes relate to survival analysis rather than standard classification trees, though specialized variants exist for that task.

Classification trees are built to assign observations to categories. They split the data in ways that group similar cases together, and each leaf of the tree ends with a label from a predefined set of categories. So, given a new observation, you follow the branches based on its feature values and land on a leaf that provides the predicted category. This is what makes them ideal for predicting categorical outcomes.

If the goal were to predict a numeric value, a regression tree would be used instead, predicting a continuous quantity. A tree can also provide class probabilities by looking at the proportion of training examples of each category in a leaf, but the final prediction for an observation is typically the category with the highest probability. Time-to-event outcomes relate to survival analysis rather than standard classification trees, though specialized variants exist for that task.