In random forest, what strategy is used to determine the outcome of the final ensemble model?

Question

Options

• ACumulative system
• BFind the average
• CVoting system
• DCumulative system

• A
  13% (3)
• B
  79% (19)
• D
  8% (2)

Explanation

Random forests combine predictions from many decision trees using two strategies depending on the task type: voting for classification (each tree "votes" for a class, and the majority wins) and averaging for regression (the numerical outputs of all trees are averaged). Both B and C are correct because a random forest applies whichever strategy matches the problem type - they are complementary, not competing, mechanisms.

Options A and D ("Cumulative system") are distractors and appear to be identical choices - neither describes a recognized ensemble aggregation method in random forests; cumulative summation without normalization is not how predictions are combined.

Memory tip: Think of a random forest as a democratic jury - for yes/no decisions, jurors vote (classification); for estimating a sentence length, they average their opinions (regression). Two strategies, one model, depending on what you're predicting.

No community discussion yet for this question.