Statistical Decision Based On The Bayesian Network Models And Applications

Bayesian network is called the probability and the combination of the graph theory, the representative of random variables and random variable represents the relationship between directly depend on the edge to form. On the one hand it using graph reveals the problem of language intuitive, on the other hand, structure according to the principle of probability problems using the complex structure, the joint probability distribution is decomposed into a series of relatively simple module, and greatly reduce the probability of the difficulty of knowledge acquisition and the complexity of the reasoning, widely used in medical diagnosis and troubleshooting, financial analysis, etc, to solve all problems and diagnostic decisions.Employment guidance problem itself is a kind of statistical decision-making problem, often have certain decision-making problem of uncertainty, the root of the problem itself mainly from the fuzziness and randomness of incomplete information and decision, not the accuracy, the limitation of human cognition and the subjective and objective understanding of the difference between the actual existence.Based on the statistics of decision making problems, according to the employment problem, Bayesian network theory, employment guidance decision-making model of employment guidance work position are analyzed, the decision-making problem of answer specific content as follows:(1) review Bayesian network theory, this paper introduces a Bayesian network structure and parameters of learning and learning process of reasoning, nets,(2) introduces statistic decision-making problem and the employment guidance and existing problems of the research significance of theory,(3) is introduced BNT toolbox of Matlab,(4) through the analysis of the employment guidance work position, and combining with actual data of decision making problems, establish decision-making Bayesian network position, and by using the model, and reasoning model of employment.Bayesian network theory papers will be applied in solving problems, and statistical decision-making problem with employment guidance for example, establish a position, a model of Bayesian decision for employment to provide the theory basis and the practical help. The process of building model with K2 structure and learning algorithm, for example, by matlab BNT toolbox, make the model of the process is more quickly, improve the general model.