Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method and device for building scorecard model based on machine learning

A machine learning and model building technology, applied in machine learning, computing models, instruments, etc., can solve problems such as low predictive ability of scorecard models and insufficient stability of machine learning models, so as to enhance audit accuracy, maintain stability and Interpretability, quality-enhancing effects

Active Publication Date: 2021-02-26
杭州排列科技有限公司
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] In order to overcome the deficiencies of the prior art, the object of the present invention is to provide a scorecard model building method and device based on machine learning, aiming to solve the problem of low predictive ability of the traditional scorecard model in the prior art and insufficient stability of the machine learning model. question

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
  • Method and device for building scorecard model based on machine learning
  • Method and device for building scorecard model based on machine learning
  • Method and device for building scorecard model based on machine learning

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment 1

[0047] The embodiment of the present invention does not limit the machine learning algorithm, which may be decision tree, random forest, gradient boosting, etc.

[0048] Such as figure 1 As shown, the embodiment of the present invention provides a method for establishing a scorecard model based on machine learning. Taking the scorecard model and decision tree as examples, the specific algorithm implementation process is as follows:

[0049] Data acquisition step S101, acquiring modeling data; the modeling data includes original one-dimensional variables;

[0050] Decision tree generation step S102, pairing the original one-dimensional variables in the modeling data to make a decision tree, generating for example figure 2 The decision tree structure in ;

[0051] Conversion step S103, for each leaf node in the decision tree, perform WOE conversion; WOE is the abbreviation of weight of evidence, meaning the weight of evidence;

[0052] Two-dimensional variable establishment ...

specific Embodiment 2

[0059] Embodiments of the present invention provide a machine learning-based scorecard model building device, including:

[0060] A derivation module, for generating derivative variables containing multidimensional information by using machine learning algorithms;

[0061] A conversion module, used for WOE conversion of derived variables;

[0062] A new module is added, which is used to put the converted new derived variables into the variable selection library of the traditional scorecard model, and use the new derived variables to establish a scorecard model.

[0063] Preferably, the machine learning algorithm is a decision tree, such as image 3 As shown, the device includes:

[0064] A data acquisition module 201, configured to acquire modeling data; the modeling data includes original one-dimensional variables;

[0065] The decision tree generation module 202 is used for pairing the original one-dimensional variables in the modeling data to make a decision tree;

[00...

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 discloses a method and device for establishing a scorecard model based on machine learning. The method includes a derivation step, a conversion step, and an addition step. A machine learning algorithm is used to generate a derivative variable containing multidimensional information. After the derivative variable is converted into WOE, when The new derived variables are put into the variable selection library of the traditional scorecard model. In this way, the stability and interpretability of the model are maintained, and the machine learning technology is used to obtain information that cannot be obtained by the traditional scorecard model. high-dimensional information. The invention can automatically generate multi-dimensional high-energy variables and automatically embed them in the establishment of the traditional score card model, thereby improving the quality of the traditional credit model, improving the correct rate of approval, and rejecting more fraudulent overdue applications.

Description

technical field [0001] The invention relates to a scorecard model of the financial credit industry, in particular to a method and device for establishing a scorecard model based on machine learning. Background technique [0002] At present, the traditional method of establishing an audit model in the financial credit industry is the traditional scorecard model, using the statistical method of logistic regression. In the newly developed Internet technology and Internet financial companies, big data machine learning and deep data mining technology have gradually become the main tools for modeling. Machine learning algorithms include decision trees, random forests, and gradient boosting machines. (Gradient Boosting Machine, GBM), Support Vector Machine (Support Vector Machine, SVM) and neural network (neural network), etc. [0003] Both the traditional scorecard model using logistic regression and the emerging machine learning model have their obvious advantages and disadvanta...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06Q10/06G06N20/00G06Q40/02
CPCG06N20/00G06Q10/06393G06Q10/067G06Q40/03
Inventor 段兆阳夏真卜象平陈薇
Owner 杭州排列科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com