Method and device for screening customer data features from large-scale feature set

The invention discloses a method and device for screening customer data features from a large-scale feature set, relates to the technical field of customer data feature screening, and can more accurately remove a large number of incomplete or insignificant features; the invention obtains more significant, complete and stable features, and optimizes the operation speed of a data feature screening link. According to the main technical scheme, the method comprises the following steps: providing screening indexes of variables based on information values, missing rates, single value rates, relevancy among features and time sequence stability; in a client data feature screening process, introducing a pre-binning operation to improve the accuracy and the stability of screening indexes (information values and time sequence stability); and reducing the complexity of a feature correlation screening operation algorithm through a preset circulation mode. When the customer data feature set has many dimensions (such as mor