Click rate estimation method and device based on subspace projection neural network
By using a click-through rate prediction method based on subspace projection neural networks, the problem of the inability to effectively integrate and extract high-order information in existing technologies is solved, enabling personalized and accurate recommendations and improving the accuracy of recommendation systems and user experience.
CN115495654BActive Publication Date: 2026-06-26BEIJING UNIV OF POSTS & TELECOMM
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING UNIV OF POSTS & TELECOMM
- Filing Date
- 2022-09-26
- Publication Date
- 2026-06-26
Smart Images
Figure CN115495654B_ABST
Abstract
The application provides a click rate estimation method based on a subspace projection neural network, and relates to the technical field of artificial intelligence, wherein the method comprises the following steps: collecting user data; performing feature processing on the user data to obtain user features, wherein the feature processing comprises normalization processing, discretization processing and unknown value processing; inputting the user features into a trained deep neural network model to output a prediction result of a user click item probability, wherein the deep neural network model is generated by training a logloss loss function; and screening an item candidate set according to the prediction result to obtain a final recommendation result. The application with the above scheme realizes end-to-end personalized and accurate recommendation for users.
Need to check novelty before this filing date? Find Prior Art