Method and device for high-speed prediction of user-financial product selection tendency based on momentum-accelerated stochastic gradient descent
A stochastic gradient descent, financial product technology, applied in finance, data processing applications, instruments, etc., to achieve the effect of good wealth distribution, good family asset allocation, and rigorous financial services
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0088] see figure 1 , figure 1 It shows the user-financial product selection tendency high-speed prediction device based on the momentum-accelerated stochastic gradient descent of the present invention, the device includes:
[0089] The data preprocessing module 510 is used to receive global user-household financial product scoring data, process the data into a data format that can be directly used in model training, and put the processed data into the data storage module 520 .
[0090] The data storage module 520 is used to store preprocessed input data, temporary variables generated during model prediction, values corresponding to initialization units, and latent feature matrices obtained from final training.
[0091] The data initialization module 530 is used to initialize the latent feature matrix for model training.
[0092] The high-speed convergence direction selection module 540 is configured to receive the initialized latent feature matrix, and determine the high-...
Embodiment 2
[0128] see figure 2 , figure 2 It shows the user-financial product selection tendency high-speed prediction method based on momentum-accelerated stochastic gradient descent of the present invention, the method includes the following steps:
[0129] S1: The server collects user-financial product rating data. The user set on the platform is denoted as M, and the product set is denoted as N. Create a matrix of |M| rows and |N| columns as the user-financial product rating matrix R. Sent to users based on momentum-accelerated stochastic gradient descent-financial product selection tendency high-speed prediction device.
[0130] S2: Use the user-financial product rating data to initialize the corresponding latent feature matrix, and construct the objective function through known rating data and corresponding predicted values.
[0131] S3: Solve the objective function according to the established objective function, and find the gradient corresponding to the decision parameter to...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


