Research on Modeling Method of Anomaly Data Detection based on Bayesian

Abstract

Today, with the help of Internet technology in the industry and consumers field, big data, cloud computing and artificial intelligence have achieving the maximum value, in data systems, both normal data in normal mode and abnormal data generated in abnormal mode are included. They determine and influence the judgment of the data user. For the identification of abnormal data, previous scholars have mainly studied it from the statistical point of view. In this paper, the idea and principle of Bayesian and Gibbs sampling algorithm are introduced. It implements parameter estimation and outliers detection simultaneously. In this way, the problem of mutual influence between parameter estimation and outliers detection is avoided.