Universal model scoring engine
AssigneePaypal, Inc.
InventorYu Liu
Life of this patent
Estimating the term…
Abstract
Methods and systems for generating a universal computer model for assessing a risk in an electronic transaction based on one or more risk assessment models are presented. The one or more risk assessment models may be incompatible with each other. Different portions of a risk assessment models may be extracted from the risk assessment models. A node structure is generated for each risk assessment model based on the portions extracted from a corresponding risk assessment model. The node structures generated based on the risk assessment models are merged to produce a merged node structure. The universal computer model is generated based on the merged node structure.
Claims20
A method comprising:
- extracting, from a first neural network configured to assess a first type of transaction risk, a first portion of computation nodes according to a first extraction scheme;
- generating a first node structure comprising a first plurality of inter-connected nodes based on the extracted first portion of computation nodes;
- extracting, from a second neural network configured to assess a second type of transaction risk, a second portion of computation nodes according to a second extraction scheme;
- generating a second node structure comprising a second plurality of inter-connected nodes based on the extracted second portion of computation nodes;
- merging the first node structure and the second node structure;
- generating a universal neural network based on the merging;
- receiving a transaction request from a user device;
- assessing a risk associated with the transaction request using the universal neural network;
- and
- processing the transaction request based on the risk.
A system comprising:
- a non-transitory memory;
- and
- one or more hardware processors coupled with the non-transitory memory and configured to read instructions from the non-transitory memory to cause the system to perform operations comprising:
- extracting, from a first neural network configured to assess a first type of transaction risk, a first portion of computation nodes using a first extraction scheme;
- generating a first node structure comprising a first plurality of inter-connected nodes based on the extracted different first portion of, computation nodes;
- extracting, from a second neural network for assessing a second type of transaction risk, a second portion of computation nodes according to a second extraction scheme;
- generating a second node structure comprising a second plurality of inter-connected nodes based on the extracted second portion of computation nodes;
- merging the first node structure and the second node structure;
- generating a universal neural network based on the merging;
- receiving a transaction request from a user device;
- assessing a risk associated with the transaction request using the universal neural network;
- and
- processing the transaction request based on the risk.
A non-transitory machine-readable medium having stored thereon machine-readable instructions executable to cause a machine to perform operations comprising:
- receiving a transaction request from a user device;
- accessing a universal neural network generating by:
- extracting, from a first neural network configured to assess a first type of transaction risk, a first portion of computation nodes according to a first extraction scheme;
- generating a first node structure comprising a first plurality of inter-connected nodes based on the first portion of computation nodes;
- extracting, from a second neural network configured to assess a second type of transaction risk, a second portion of computation nodes according to a second extraction scheme;
- generating a second node structure comprising a second plurality of inter-connected connected nodes based on the second portion of computation nodes;
- and
- merging the first node structure and the second node structure;
- assessing a risk associated with the transaction request using the universal neural network;
- and
- processing the transaction request based on the risk.
Dependencies are inferred from the claim text.