Time-Based Ensemble Learning Patent Granted to Visa International

ChangeBridge: Patent Grants - AI & Computing (G06N)

Published April 7th, 2026

Detected April 7th, 2026

Summary

The USPTO granted Patent US12596964B2 to Visa International Service Association covering systems and methods for time-based ensemble learning using both supervised and unsupervised machine learning models. The patent includes 12 claims and covers techniques for augmenting training data with time delays and combining multiple ML model outputs into an ensemble. The patent names Yinhe Cheng, Yu Gu, and Sam Peter Hamilton as inventors.

View original document View source feed page

What changed

The USPTO officially granted patent rights to Visa International Service Association for time-based ensemble learning technology combining supervised and unsupervised machine learning models. The patent covers systems that receive time series training data, apply time delay augmentation to data instances, train multiple supervised and unsupervised models on the augmented dataset, and generate ensemble outputs from combined model predictions.

For competitors and technology developers, this grant establishes enforceable IP rights that could restrict development of similar ensemble learning systems using time-delay augmentation. Visa gains exclusive rights to practice the claimed inventions for payment processing optimization, fraud detection, transaction analysis, or other financial services applications. Companies developing machine learning systems in fintech, payments, or related fields should conduct freedom-to-operate analyses to assess infringement risk.

What to do next

Monitor for competitive landscape changes in AI/ML patent portfolios
Review potential licensing implications if developing similar ensemble learning systems
Assess whether existing AI/machine learning processes may infringe on the granted claims

Source document (simplified)

← USPTO Patent Grants

System, method, and computer program product for time-based ensemble learning using supervised and unsupervised machine learning models

Grant US12596964B2 Kind: B2 Apr 07, 2026

Assignee

Visa International Service Association

Inventors

Yinhe Cheng, Yu Gu, Sam Peter Hamilton

Abstract

Provided are systems for ensemble learning with machine learning models that include a processor to receive a training dataset of a plurality of data instances, wherein each data instance comprises a time series of data points, add an amount of time delay to one or more data instances to provide an augmented training dataset, select a first plurality of supervised machine learning models, select a second plurality of unsupervised machine learning models, train the first plurality of supervised machine learning models and the second plurality of unsupervised machine learning models based on the augmented training dataset, generate an ensemble machine learning model based on outputs of the supervised machine learning models and unsupervised machine learning models, and generate a runtime output of the ensemble machine learning model based on a runtime input to the ensemble machine learning model. Methods and computer program products are also provided.