Java: Data Science Made Easy

Using a Bayesian network

To model this type of network using Java, we will create a network using JBayes (https://github.com/vangj/jbayes). JBayes is an open source library for creating a simple Bayesian Belief Network (BBN). It is available at no cost for personal or commercial use. In our next example, we will perform approximate inference, a technique considered less accurate but allowing for decreased computational time. This technique is often used when working with big data because it produces a reliable model in a reasonable amount of time. We conduct approximate inference by using weight sampling of each node. JBayes also provides support for exact inference. Exact inference is most often used with smaller datasets or in situations ...