Method, apparatus, and computer program for selecting generated samples

What is AI technical title?
AI technical title is built by Patsnap AI team. It summarizes the technical point description of the patent document.
By grouping and selecting generated samples based on their contribution to the learning target model, the method enhances the effectiveness of machine learning by improving model performance and reducing the impact of non-contributory or degrading data.

JP2026096941APending Publication Date: 2026-06-15NTT DOCOMO INC

0 Cites 0 Cited by

Patent Information

Application Number: JP2025203301
Authority / Receiving Office: JP · JP
Patent Type: Applications
Current Assignee / Owner
Priority Date: 2024-12-03
Filing Date: 2025-11-26
Publication Date: 2026-06-15

Application Information

Patent Timeline

03 Dec 2024

Priority

26 Nov 2025

Application

JP2025203301

15 Jun 2026

Publication

JP2026096941A

IPC: G06F18/23

Get more patent work done with AI

Run novelty searches—or use them for FTO, design risk, drafting and technical problem-solving.

AI Tagging

Technology Topics

Algorithm Computer engineering

Explore More Agents

Novelty Search
Search existing technologies and assess novelty
↗
FTO
Analyze whether a product may infringe others' patents
↗
Design FTO
Check prior-design risk for exterior design
↗
Drafting
Draft patent application text based on a technical solution
↗
Find Solutions with TRIZ
Generate feasible solution to solve your technical challenge
↗
IP MCP Servers
Connect trusted patent search and cited evidence to Claude, Cursor or your own AI workflow
↗

AI Technical Summary

Technical Problem

The newly generated data in machine learning is not always effective for model training, often containing non-contributory or degrading data, making it challenging to select effective generated samples for downstream model training.

Method used

A method is provided to divide a sample set into groups, cluster samples, and select generated samples based on their contribution to a learning target model, using a genetic algorithm as a fitness supervisor function to identify and prioritize samples that enhance model performance.

Benefits of technology

This approach improves the performance of downstream models by selecting generated samples that contribute effectively to training, reducing overfitting and enhancing generalization.

✦ Generated by Eureka AI based on patent content.

Patent Text Reader

Abstract

This invention provides a method, apparatus, and computer program product for selecting generated samples to improve the performance of downstream models. [Solution] A method for selecting generated samples includes dividing a sample set, which includes real samples and generated samples, into multiple groups, and selecting generated samples within the sample set by selecting a portion of the generated samples within each of the multiple groups based on the contribution of the generated samples in each group to the model to be trained.

Need to check novelty before this filing date? Find Prior Art