Search code examples
pythonmachine-learningartificial-intelligenceprobabilitynaivebayes

Naive Bayes same probability


I am working on Bayes and Naive Bayes theorem. I am stuck in the condition when the probabilities are the same for both labels. How does Naive Bayes handle this scenario?


Solution

  • In case of equality you may arbitrarily set the sample to any class, exactly for that reason: the sample can belong to any of the classes with the same probability, no preference.

    Now you only need to select the strategy: you can choose constantly one specific class, use apriory class distribution, choose in random or using round-robin approach. Most simple is to choose randomly.