Federal learning method and system for e-commerce platform

The invention relates to a federal learning method and system for an e-commerce platform, and relates to the technical field of data security and privacy protection, and the method comprises the steps: generating a non-independent identically distributed data model according to the type of a client, and obtaining a skewness value of the client; selecting a plurality of clients from the client pool as training data providers according to the meta-parameters; and evaluating an output result of the federal learning model, and feeding back and adjusting meta parameters until a model evaluation result reaches an expectation. According to the method, the data isomerism of the client is deduced and privacy protection is performed by constructing the non-independent identically distributed data model aiming at the data characteristics of the e-commerce platform, and the quality of the training data is improved by introducing the meta-parameter balance federated learning model to train the required data volume and the