Description

In data mining once we construct a Bayesian network is defined, the next thing to do is to quantify the relationships between connected nodes. This is done by specifying a conditional probability distribution for each node or else to build the conditional probability table (CPT). In a Bayesian network every node must have a CPT, associated with it. Conditional probabilities represent likelihoods based on prior information or past experience. In Existing Method was developed on Bayesian Network, used to model uncertainties between only discrete variables. It was lack in to handle the continuous variables and also time complexity was increase and also it was lack in accuracy. In our proposed, we built a Hybrid Bayesian Network to model the uncertainties between both discrete and continuous variable using dynamic discretization algorithm. Identify the bigamy node in BN using Binary Factorization Algorithm. Add that detected node and their parent to the nodes of the bigamy node. And then add this bigamy node to bigamy network, finding the bigamy node to reduce the time complexity. Convert the bigamy node into binary tree using binary factorization algorithm. CPT size will be optimized; we can find the probability of certain event depending on that result construct the new Hybrid Bayesian network.