Random Forest Interchange Code Prediction for Low Latency Processing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current credit card transaction processing systems lack the ability to accurately predict interchange fees in real-time, leading to undesirable latency and significant revenue impact for merchants, as these fees are complex and determined only after transactions are processed.

Innovation Solution

A computer-implemented method using a random forest model to predict interchange codes by transforming historical transaction data into BIN features and transaction features, generating predicted interchange code probability distributions, and calculating weighted sum interchange charges, which are then used to create a lookup table for timely and accurate fee assessment during transaction processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional interchange fee assessment methods are used, then accurate fee calculation is achieved, but processing latency increases and real-time prediction is not possible

Engineering Contradiction:

Improveinterchange fee prediction accuracyVSAvoidtransaction processing latency

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary actions by training a random forest model offline using historical transaction data, BIN features, and interchange codes. The trained model and lookup tables are prepared in advance, enabling rapid real-time predictions during actual transaction processing without incurring latency

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system creates simplified copies of the complex interchange fee assessment process by generating lookup tables that map BIN features and transaction characteristics to predicted interchange codes and fees. These lookup tables serve as lightweight proxies that can be queried rapidly during transaction processing

Inventive Principle:

Principle #26Copying

2Reliability

If interchange fees are assessed only after transaction processing, then complex fee structures are handled accurately, but merchants face significant end-of-month financial surprises

Engineering Contradiction:

Improvefee assessment accuracyVSAvoidreal-time fee visibility

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The system implements feedback by comparing predicted interchange fees with actual assessed fees, using the differences to continuously improve the model's accuracy. This feedback loop enables the system to maintain reliable predictions while providing real-time fee visibility to merchants

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs preliminary fee assessment by predicting interchange codes and calculating fees before transactions are fully processed. This preliminary assessment provides merchants with real-time visibility into expected costs, allowing for better cash flow management while maintaining accurate final billing

Inventive Principle:

Principle #10Preliminary action

3Productivity

If simple interchange fee estimation is used, then processing speed is maintained, but prediction accuracy is insufficient for complex fee structures

Engineering Contradiction:

Improvetransaction processing speedVSAvoidinterchange code prediction accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system creates simplified lookup tables that copy the essential fee assessment logic in a query-optimized format. These lookup tables enable fast retrieval of predicted interchange codes and fees based on BIN features and transaction characteristics, maintaining high processing speed

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The system transforms BIN identifiers into numerical features and categorizes transaction characteristics into discrete parameters. By converting complex input data into standardized numerical representations, the random forest model can efficiently process transactions while maintaining high prediction accuracy for interchange codes

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11775969B2Low latency bank card type prediction system for estimation of interchange codes during transaction processing

Publication Date: 2023.10.03 TOAST INC

AI summary

A computer-implemented method for predicting interchange charges includes: retrieving a historical transactions set, where each completed transaction in the set includes transaction features, a bank identification number (BIN), and a corresponding true interchange code; transforming all BINs in the set into a corresponding plurality of BIN features that comprise probabilities; creating a first training set including all transaction features, all pluralities of BIN features, and all true interchange codes associated with the historical transactions set; training a random forest model using the first training set and generating a second training set including rounded BIN features, rounded transaction features, discrete ones of the transaction features, and the true interchange codes; training the random forest model using the second training set to generate a trained random forest model for prediction of the interchange codes; and executing the trained random forest model for new transactions to generate corresponding predicted interchange codes.