In a decision tree, the root node represents the first predictive independent variable.

• A. The first predictive independent variable
• B. The last predictor
• C. The target attribute
• D. A random attribute

Answer: (A)

In a decision tree the root node is where the first split happens, based on the predictor that provides the most informative division of the data given the chosen impurity measure (like information gain or Gini). This makes it the first predictive independent variable used to partition the data, guiding the initial direction of the tree. The target attribute isn’t used to split at the root—it's what you’re trying to predict, so it sits as the outcome node rather than a starting splitter. A random attribute wouldn’t be selected because it wouldn’t maximize the information gained from the split.