Data analysis device, display device, data analysis method and program

JPWO2024157389A5Active Publication Date: 2025-09-12NEC CORP
0 Cites 0 Cited by

Patent Information

Application Number
JP2024572599
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2025-07-02
Publication Date
2025-09-12
Estimated Expiration
2043-01-25

AI Technical Summary

Technical Problem

Current reinforcement learning methods lack the ability to effectively evaluate the influence of individual data on learning results, which can lead to suboptimal performance and inefficient data usage, particularly in scenarios where data age and environmental conditions impact learning outcomes.

Method used

A data analysis device and method that updates policy and value functions using evaluation functions, allowing for the evaluation of data influence by performing reinforcement learning with and without specific data, and displaying the degree of influence for data in the training dataset, enabling data selection and environment adjustments for improved performance.

Benefits of technology

Enables the evaluation and optimization of data influence on reinforcement learning results, leading to improved robot performance, reduced data set size, and faster convergence times by prioritizing new data and adjusting environmental settings based on evaluated influences.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

This data analysis device is provided with: a learning means that performs reinforcement learning in which a policy function is updated using an evaluation function that includes a value function, and the value function is updated using an evaluation function that includes the policy function; and an evaluation means that evaluates the degree of influence imposed by evaluation data on learning results using an evaluation of learning results obtained when the reinforcement learning is performed using a data set that includes the evaluation data, and an evaluation of learning results obtained when the reinforcement learning is performed using a data set that does not include the evaluation data.
Need to check novelty before this filing date? Find Prior Art