Definition of Naive Bayes Classifier A naive Bayes classifier is a machine learning algorithm that relies on the assumption that the features of a given object are independent of each other. This algorithm is often used for text classification tasks, such as sorting email into spam and non-spam folders.