Neural Network Fraud Detection Variable Update

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Traditional fraud detection systems for transactions face challenges in updating fraud detection variables frequently due to the manual and time-consuming process, leading to suboptimal fraud detection effectiveness.

Innovation Solution

An automated system using a neural network that continuously analyzes transaction data to update fraud detection variables by cycling through consumer and merchant transaction histories, adjusting weights to improve prediction accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If fraud detection variables are updated manually, then the updating process is simple to implement, but the updating frequency is low and fraud detection effectiveness is reduced

Engineering Contradiction:

Improveupdating frequencyVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system enables fraud detection variables to update themselves automatically through a neural network that continuously processes transaction data. The neural network self-adjusts weights and generates updated fraud detection variables without human intervention, allowing frequent updates while maintaining operational simplicity through automation.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The manual mechanical process of updating fraud detection variables is replaced with an automated neural network system. The neural network uses machine learning algorithms to automatically analyze transaction patterns and generate updated fraud detection variables, substituting human manual operations with an intelligent automated system.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If fraud detection variables are updated frequently, then fraud detection effectiveness is improved, but the manual updating process becomes more time-consuming and labor-intensive

Engineering Contradiction:

Improvefraud detection effectivenessVSAvoidupdating time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The neural network operates continuously to analyze transaction data and update fraud detection variables without interruption. This continuous operation ensures that fraud detection variables are constantly refined based on the latest transaction patterns, improving detection effectiveness while eliminating the time loss associated with periodic manual updates.

Inventive Principle:

Principle #20Continuity of useful action

Solution Approach 2:

The system implements a feedback mechanism where the neural network continuously processes transaction data, compares predictions with actual outcomes, and automatically adjusts fraud detection variables based on this feedback. This closed-loop system improves fraud detection effectiveness over time while requiring no additional manual time investment.

Inventive Principle:

Principle #23Feedback

3Adaptability or versatility

If manual updating of fraud detection variables is performed, then the process is easy to implement, but the variables do not reflect current transaction history

Engineering Contradiction:

Improvereflection of current transaction informationVSAvoidupdating effort

Core Design Contradiction:

Adaptability or versatilityVSEase of operation

Solution Approach 1:

The neural network is pre-configured with the capability to process transaction data and generate fraud detection variables. It continuously performs preliminary analysis of transaction patterns before fraud incidents occur, ensuring that fraud detection variables always reflect the most current transaction history without requiring manual intervention.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system automatically updates fraud detection variables by itself using the neural network to process new transaction data. This self-service capability ensures that fraud detection variables continuously adapt to reflect current transaction patterns without requiring manual updating effort, maintaining both ease of operation and adaptability.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS10956987B2Applying multi-dimensional variables to determine fraud

Publication Date: 2021.03.23 AMERICAN EXPRESS TRAVEL RELATED SERVICES CO INC

AI summary

The systems and methods herein may include receiving a plurality of transactions for a plurality of consumers, wherein each respective transaction of the plurality of transactions is between a consumer of the plurality of consumers and a merchant of a plurality of merchants; automatically inputting the plurality of transactions into a neural network; automatically analyzing the plurality of transactions over a plurality of iterations, wherein an iteration of the plurality of iterations comprises cycling through a consumer transaction history associated with the consumer, wherein the consumer transaction history has a consumer transaction sequence associated with the consumer; and automatically updating over the plurality of iterations, a previous fraud detection variable associated with the consumer and/or the merchant to generate updated fraud detection variables, in response to the analyzing the plurality of transactions.