Prediction method and device for scoring card model

A prediction method and scorecard technology, which is applied in the computer field, can solve the problems that the scorecard model cannot be mapped into, and the prediction of the multi-party scorecard model that does not meet the privacy protection requirements, etc.

Active Publication Date: 2021-05-18
BEIJING REALAI TECH CO LTD
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, WeBank's FATE solution is aimed at a relatively simple forecasting scenario, which does not meet the needs of privacy-preserving multi-party scorecard model forecasting
The secret sharing scheme provides a solution to split the data into "secrets" and perform secure calculations according to the protocol. At present, it is not possible to map the scorecard model with logical information into "secrets" and ensure the privacy protection of multi-party information

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Prediction method and device for scoring card model
  • Prediction method and device for scoring card model
  • Prediction method and device for scoring card model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0031] refer to figure 2 The flow chart of the prediction method of the scorecard model shown, the method includes the following steps:

[0032] Step S202, acquiring a scorecard model; wherein, the scorecard model includes an initial prediction algorithm for predicting objects to be scored.

[0033] In this embodiment, the scorecard model is a scorecard model involving multiple participants, one of which is the model provider of the scorecard model, and the data used in the process of predicting the object to be scored through the scorecard model is Provided by at least two other parties (that is, data holders) other than the model provider. Taking the scorecard model involving three parties (a, b, and c) as an example, as shown in Table 1, in the scenario of the scorecard model involving three parties, the final score of the object to be scored is obtained by the model holder and is obtained by the model holder. get.

[0034]The scorecard model can include the following s...

Embodiment 2

[0056] refer to image 3 The structural block diagram of the forecasting device of the scorecard model shown in the present disclosure, the predicting device of the scorecard model provided by the embodiment of the present disclosure includes the following modules:

[0057] A model acquiring module 302, configured to acquire a scorecard model; wherein, the scorecard model includes an initial prediction algorithm for predicting objects to be scored;

[0058] A conversion module 304, configured to convert the initial prediction algorithm into a target prediction algorithm based on a preset conversion algorithm; wherein, the intermediate calculation results of the conversion algorithm are stored in different participants in a secret manner;

[0059] The prediction module 306 is configured to predict the score value of the object to be scored according to the scorecard model using the target prediction algorithm, and the score value is obtained by the model provider.

[0060] The...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention relates to a prediction method and device for a scoring card model. The method comprises the following steps: acquiring the scoring card model, wherein the scoring card model comprises an initial prediction algorithm for predicting the to-be-scored object; converting the initial prediction algorithm into a target prediction algorithm based on a preset conversion algorithm, wherein an intermediate calculation result of the conversion algorithm is stored in different participants in a secret mode; and according to the scoring card model using the target prediction algorithm, predicting a scoring value of the to-be-scored object, and obtaining the scoring value by a model provider. According to the method and the device, data privacy protection can be well realized in a multi-party scoring card model prediction scene.

Description

technical field [0001] The present disclosure relates to the field of computer technology, in particular to a method and device for predicting a scorecard model. Background technique [0002] In the scenario of a multi-party scorecard model (taking three-party participation as an example), two of the participants each hold a part of the feature data, called the data holder, and the third participant holds the scorecard model, called the model holder. Proper. In the score prediction process, the final score is obtained by the model holder, but the feature data used to calculate the score comes from the data holder. In the data privacy protection scenario, the following requirements need to be met: the characteristic data of the data holder cannot be directly or indirectly leaked to other participants, and the score model of the model holder cannot be leaked to other participants; at the same time, the final score collection On the model holder side. [0003] In order to me...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06F21/62G06F21/60
CPCG06F21/602G06F21/6245
Inventor 不公告发明人
Owner BEIJING REALAI TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products