Which statement best describes Naive Bayes in predicting categorical outcomes?

• A. It Uses Variance to Estimate Class Membership.
• B. It Cannot Handle Categorical Outcomes.
• C. It Uses Probability to Estimate Class Membership.
• D. It Ignores Probabilities and Relies on Mean Values

Answer: (C)

Naive Bayes is a probabilistic classifier that uses Bayes’ theorem to estimate how likely each possible class is for a given instance. For categorical outcomes, it computes the posterior probability P(class | features) for every class and then picks the class with the highest probability. This relies on the prior probability of each class, P(class), and the likelihoods P(feature_i | class) for each feature, multiplied together under the assumption that features are conditionally independent given the class. This independence assumption keeps the calculation simple even with many features and lets the model handle categorical data by choosing appropriate likelihoods (such as Bernoulli for binary features or multinomial for count-based features). In short, it predicts category by weighing evidence with probabilities and selecting the most probable category.