What is the relationship between Naïve Bayes and logistic regression in terms of decision boundaries?

Naïve Bayes and logistic regression often produce similar linear decision boundaries but differ in modeling assumptions.

Naïve Bayes is a generative model, while logistic regression is discriminative. Under certain assumptions (feature independence and correct distribution), both yield linear decision boundaries in log-space. However, logistic regression directly optimizes conditional likelihood, making it more flexible and often more accurate with large datasets.