Data processing method and device
The data processing method and apparatus address the challenge of lengthy estimation times and high computational load by creating regression models and focusing on interpolation data within the data range of explanatory variables, facilitating efficient material formulation searches.
JP2025102161APending Publication Date: 2025-07-08PROTERIAL LTD
Patent Information
- Application Number
- JP2023219437
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-12-26
- Publication Date
- 2025-07-08
AI Technical Summary
Technical Problem
The large number of combinations of material formulations in composite materials leads to prolonged estimation times and increased computational load in searching for formulations with desired properties.
Method used
A data processing method and apparatus that create a regression model using learning data, generate candidate data through random settings or predetermined rules, and exclude extrapolation data to focus on interpolation data within the data range of the explanatory variables, applying this data to the model for estimation.
Benefits of technology
This approach reduces estimation time and computational load by focusing on interpolation data, enabling efficient searching for material formulations with desired characteristics.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure 2025102161000001_ABST
Abstract
To provide a data processing method and device which is suitable to retrieve a material combination, shortens an estimation time and reduces a calculation load.SOLUTION: A data processing method includes: a model creation step for creating a regression model 32 showing a correlation between an explanatory variable and an objective variable by using learning data 31 including a plurality of data of the explanatory variable and the objective variable; a candidate data creation step for creating estimation source candidate data 33 including the plurality of data by setting data of each parameter of the explanatory variable at random or applying a prescribed rule set in advance to create data; an estimation source data generation step for generating estimation source data 34 on the basis of interpolation data in the estimation source candidate data 33, being the interpolation data existing within the range of the explanatory variable included in the learning data 31; and an estimation step for respectively applying the data included in the estimation source data 34 to the regression model 32 to respectively calculate estimates of the objective variable.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art
Citation Information
Patent Citations
Learning model generation method, program, storage medium, and learned model
JP2021183666A