User portrait recognition model training method and device, readable storage medium and product

Abstract

The invention discloses a user portrait recognition model training method and device, a readable storage medium and a product, and relates to deep learning and big data in data processing. According to the specific implementation scheme, the method comprises the following steps: acquiring multiple groups of user portrait data and label information corresponding to the user portrait data; performing training operation on a preset reference network model and a shadow network model through the multiple groups of user portrait data and the label information to obtain a first prediction label output by the reference network model and a second prediction label output by the shadow network model; updating the training parameters of the reference network model and the shadow network model according to the label information, the first prediction label and the second prediction label to obtain a first training parameter corresponding to the reference network model and a second training parametercorresponding to the shadow network model; and training the reference network model and the shadow network model by using the first training parameter and the second training parameter. Therefore, the robustness and the recognition precision of the user portrait recognition model can be improved.

Description

technical field [0001] This application relates to deep learning and big data in data processing, and in particular to a user portrait recognition model training method, device, readable storage medium and product. Background technique [0002] User portraits are also called user roles. In the context of the big data era, user information is flooded in the network, and each specific information of users is abstracted into labels, and these labels are used to concretize user images, so as to effectively provide users with targeted information. sexual services. [0003] In order to realize the recognition operation of user portraits, in the prior art, a large amount of pre-collected user portrait training data with label information is generally used, and user portrait data are divided into multiple groups to perform training operations on preset network models. [0004] However, in the process of model training using the above method, due to the high-dimensional and sparse f...

AI Technical Summary

Problems solved by technology

[0004] However, in the process of model training using the above method, due to the high-dimensional and sparse features in the label recognition scene of user portraits, the solution space of the model is complex, which makes the model more sensitive and fragile

If there is a set of poor quality training data, it may lead to overfitting of the model, which in turn leads to low recognition accuracy of the trained model

Images