CPC G06N 20/20 (2019.01) [G06Q 40/03 (2023.01)] | 19 Claims |
1. A method for generating a machine learning model, the method implemented by one or more computers and comprising:
training a machine learning model by applying a subset of credit data as training input to the machine learning model, thereby providing a first trained machine learning model, wherein the subset of the credit data is selected from a credit data sample based on sampling the credit data according to a first user input data characterizing a product or service and the credit data sample comprises third party data;
re-training the first trained machine learning model by applying at least third user input data as training input to the first trained machine learning model, thereby generating a second trained machine learning model, wherein the third user input data is obtained based on sampling criteria and comprises first party data comprising credit reports or applications of actual customers associated with the product or service;
ensembling the first and second trained machine learning models to thereby combine the first and second trained machine learning models into a refined version of the first trained machine learning model;
deploying the refined version of the first trained machine learning model in a computing environment as a web service having an associated application programming interface (API) endpoint; and
executing the refined version of the first trained machine learning model to generate and return credit scores or decisions with respect to the product or service based on input client data received via the API endpoint from a loan origination system.
|