Federated Machine Learning Model Distribution for Secure Transaction Authentication

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing security measures for computer systems, particularly in unsecured networks, rely on vulnerable single-factor authentication methods that are insufficient for modern data protection regulations, and the use of machine learning algorithms raises concerns about sensitive information privacy and transmission.

Innovation Solution

Implementing a federated machine learning model that distributes portions across user devices and edge servers to evaluate transaction requests locally, generating scores instead of transmitting sensitive information, thereby enhancing security and compliance with privacy regulations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If machine learning algorithms are used to identify patterns in requests, then security detection capability is improved, but sensitive information privacy is compromised due to transmission requirements

Engineering Contradiction:

Improvesecurity detection capabilityVSAvoidsensitive information privacy

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The machine learning model is segmented into multiple portions distributed across different systems. The first portion resides on the server system and the second portion resides on the user device. This segmentation allows the model to process requests locally without transmitting sensitive information, resolving the contradiction between security detection capability and information privacy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

A federated machine learning architecture acts as an intermediary between the server system and user device. This intermediary enables collaborative pattern recognition for security purposes while maintaining data locality, allowing security detection to improve without compromising sensitive information privacy through the use of distributed model portions.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Device complexity

If single-factor authentication methods are used, then system complexity is reduced, but security adequacy deteriorates under modern data protection regulations

Engineering Contradiction:

Improveauthentication system complexityVSAvoidsecurity adequacy

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The federated machine learning model provides multi-functionality by simultaneously performing pattern recognition for security assessment and enabling distributed computation. This allows the system to achieve multi-factor authentication capabilities without proportionally increasing system complexity, as the same distributed architecture serves multiple security functions.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS12081541B2Device-side federated machine learning computer system architecture

Publication Date: 2024.09.03 PAYPAL INC

AI summary

Techniques are disclosed in which a computer system receives a transaction request and uses a federated machine learning model to analyze the transaction request. A server computer system may generate a federated machine learning model and distribute portions of the federated machine learning models to other components of the computer system including a user device and/or edge servers. In various embodiments, various components of the computer system apply transaction request evaluation factors to the portions of the federated machine learning model to generate scores. The server computer system uses the scores to determine a response to the transaction request.