Differentially Private Naïve Bayes Classifier Using Smooth Sensitivity

Abstract There is increasing awareness of the need to protect individual privacy in training data used develop machine learning models. Differential Privacy a strong concept protecting individuals. Naïve Bayes popular algorithm, as baseline for many tasks. In this work, we have provided differentially private classifier that adds noise proportional smooth sensitivity its parameters. We compare our results Vaidya, Shafiq, Basu, and Hong [1] which scales global Our experimental on real-world datasets show significantly improves accuracy while still guaranteeing ? -differential privacy.