Machine learning engine research method

The invention discloses a machine learning engine research method, and the method comprises the following steps: S1, providing a modeling framework based on an interactive learning technology for a user, and supporting a graphical mode for modeling mining; s2, integrating more than three machine learning frameworks such as Spark, and setting various algorithms such as classification, regression, clustering and association rules; s3, enabling deep learningto be a method based on data representation learning in machine learning; s4, based on the operation data of the company information system,researching the application of machine learning and deep learning algorithms in the aspects of information system operation monitoring, information equipment fault diagnosis, system security risk prediction, information user behavior analysis prediction and the like. According to the method, the technical threshold of modeling programming is reduced, so that a company service department can easilyobtain the model training