Bayesian networks are a new tool in modeling static and dynamic phenomena and systems and are used in various fields such as disease diagnosis, weather forecasting, decision making and classification. A Bayesian network is a probabilistic graph model that represents the cause-and-effect relationships between random variables and consists of a directed acyclic graph and a set of conditional probabilities. Two important issues in modeling a dataset with a Bayesian network are structural learning and network parametric learning. In this paper, we consider a Bayesian network with a known structure and try to learn the network structure using two common algorithms, PC and $ K_{2} $ , through simulation. Then, we learn the network parameters and obtain maximum likelihood, maximum posterior likelihood, and posterior mean estimates of the parameters of interest. Next, we compare the performance of the estimates using the Kullback-Leibler divergence criterion and finally, using a real dataset, we learn the network's structure and parameters to demonstrate the feasibility of implementing the proposed methods on real data.