Decision Trees DATA-SCIENCE-FOUNDATIONS Practice Questions (Page 1)

Questions 1–5

1. 1application · medium

   A data scientist has trained a decision tree that achieves 100% accuracy on the training set but only 65% on the validation set. To improve generalization, they decide to use cost-complexity pruning. What does this technique do?

   • AIt replaces the tree with a random forest.
   • BIt removes the least important features before training.
   • CIt grows the tree to full depth and then removes branches that provide little predictive power relative to their complexity.
   • DIt stops the tree from growing when the node has fewer than a certain number of samples.
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   Which decision tree algorithm is known for producing binary trees and using Gini impurity for classification and variance reduction for regression?

   • AID3
   • BCHAID
   • CCART
   • DC4.5
   Reveal Solution Discussion Select an answer first
4. 3foundation · easy

   Which component of a decision tree is a terminal node that holds the final prediction or class label?

   • ALeaf node
   • BBranch
   • CRoot node
   • DInternal node
   Reveal Solution Discussion Select an answer first
5. 4application · easy

   A data scientist is using the ID3 algorithm to build a decision tree. The dataset has only categorical features and no missing values. Which characteristic of ID3 is most relevant to this situation?

   • AIt uses information gain to select the best attribute for splitting.
   • BIt can handle numerical features directly.
   • CIt supports multi-way splits on numerical features.
   • DIt handles missing values automatically.
   Reveal Solution Discussion Select an answer first
6. 5application · medium

   A data analyst is building a decision tree to predict customer satisfaction. The dataset has a numerical feature 'age' and a categorical feature 'region' with values 'North', 'South', 'East', 'West'. The analyst wants to use both features. Which approach is most appropriate for the 'region' feature in a CART tree?

   • AAssign numerical codes 1, 2, 3, 4 to the regions.
   • BDrop the 'region' feature because decision trees cannot handle categorical features.
   • CConvert 'region' to a binary variable for each category (one-hot encoding) and use them as separate features.
   • DUse the 'region' feature directly, allowing the tree to split on subsets of categories.
   Reveal Solution Discussion Select an answer first