Data lake-based store entering rate prediction method and system of brand store

The invention discloses a brand store entry rate prediction method and system based on a data lake, and the method specifically comprises the steps: collecting offline data needed by model training, carrying out the centralized storage of the offline data in the data lake, carrying out the data preprocessing of the collected data, setting a prediction target dimension, constructing a sample and features according to the dimension, and carrying out the prediction of the entry rate of a brand store. The optimal model and parameters are found through automatic modeling and parameter searching of the model, then all samples are input into the optimal model to obtain a final to-be-online model, and the final to-be-online model is used for online prediction of the store entering rate of the to-be-predicted samples and estimation of the long-term and short-term store entering rates of brand stores at the to-be-opened position, so that a foundation is laid for further estimation of store opening earnings. According t