Pengembangan Model Bayesian Network

Abstract

Bayesian Network (Bayesian Network) is a directed acyclic graph that used to express the dependence of the probability of causation between variables that focus on knowledge representation and inference uncertainty, it is regarded as a raw cognitive model in uncertainty reasoning based on probability. Bayesian networks which used probabilistic graph model (probabilistic graphical models) is abbreviated as PGM. Decency in Bayesian Network is located in a graph that illustrates the results of the various forms of probability variables and criteria referred to the model. Employee performance is usually influenced by the ability of the environment inside and outside the company / institution. Performance is taken from a variety of information and is stored in the form of employee data. Employee data in structural stacking with the development of Bayesian Network is more efficient and easier to conclude that a good performance is taken through several variables with multiple parameters. By using the theory of Probabilistic Graphical Models and systematics (PGM) from Bayesian Network produced four models and four graphs of the performance with the addition of the weight value of each character /Maasing parameters on each variable. Therefore the techniques needed for the learning of Bayesian networks to implement and develop a Bayesian Network models for analyzing the performance of employees. This causes a very good Bayesian networks are applied in various fields.