What are the Pros/Cons of Naive Bayes?
Pros: Computational efficiency
Cons:Independence assumption is not realistic for many data sets
Pros: Computational efficiency
Cons:Independence assumption is not realistic for many data sets
Gaussian Naive Bayes accounts for continuous features by calculating the conditional likelihood rather than conditional probability of an observation belonging to each class
If a feature appears zero times within a particular class, the computed likelihood score for an observation belonging to that class will be zero
Naive Bayes uses the framework of Bayes Theorem and the assumption of conditional independence between all pairs of predictors
In some classification contexts, it might be more of interest to obtain predicted probabilities of class membership rather than simply the labels themselves.
Logistic regression is the most traditional classification algorithm and preserves many of the advantages in interpretation as linear regression for a continuous outcome.
Partner Ad