Decomposable Naive Bayes Classifer for Distributed Data using Directed Acyclic Graph

A common constraint in distributed data is that the database cannot be moved to other network sites due to computational costs, data size, or privacy considerations. All of the existing distributed algorithms for classifying data using Naive Bayes are designed for horizontally distributed or special case of vertically distributed data where different sites contain different attributes for a common set of entities. In this paper, we present a framework including a general model and a decomposable version of Naive Bayes Classifier using directed acyclic graph (NBC-DAG) in d-dimensional space across vertically distributed data in the most general situation in which existing distributed databases. The goal of our algorithm is to minimize the cost of communication among the database nodes by gathering statistical summaries at each distributed database and then passing messages describing those summaries between the participating sites which preserves the privacy of the data.