Which of the following is a probabilistic classifier among the listed options?

• A. Neural networks
• B. Decision trees
• C. Random forests
• D. Naive Bayes

Answer: (D)

A probabilistic classifier is one that provides explicit probabilities for each class given the input. Naive Bayes does this by applying Bayes' theorem to compute the posterior P(class|features), combining the prior for each class with the likelihood of the observed features, and it reports those probabilities for decision-making. It assumes features are conditionally independent given the class, which makes the calculation straightforward as a product of individual feature likelihoods. The other models shown can produce probabilities or scores, but they do not embody a formal probabilistic model of the data in the same way: neural networks yield class predictions (often with softmax-provided probabilities) without modeling P(features|class) in a generative sense, while decision trees and random forests focus on splits and votes, giving labels or vote-based probabilities rather than explicit posterior probabilities derived from Bayes’ rule.