Consumable Deterioration Graph Display for Prediction Confidence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face difficulties in determining the accuracy of prediction data when using machine learning models for predicting the deterioration of consumables, such as batteries, as existing methods lack clear visualization and comparison tools.
Innovation Solution
A prediction data display device and method that uses a trained Gaussian process regression model to generate and display graphs of predicted and historical data, with display appearances determined by similarity, including attenuation factors and confidence intervals, allowing users to visually assess the accuracy of predictions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning models are used to predict deterioration of consumables, then prediction capability is improved, but ease of determining prediction accuracy deteriorates
Solution Approach 1:
The patent creates visual copies of historical data graphs and training data graphs that can be directly compared with prediction results. By displaying multiple graphs showing characteristic data at different time points alongside prediction data, users can visually verify accuracy without complex analysis, making the determination of prediction correctness intuitive and easy.
Solution Approach 2:
The patent uses simple visual display elements (graphs and charts) rather than complex analytical tools to assess prediction accuracy. These visual representations provide immediate, intuitive feedback about prediction correctness without requiring expensive or complicated verification systems.
2Measurement precision
If detailed prediction data is displayed, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent segments the display into multiple distinct graph components, each showing specific aspects of the data (historical characteristic data, training data, prediction results). This segmentation allows detailed information to be presented in an organized, manageable way that doesn't overwhelm the user or require complex integration systems.
Solution Approach 2:
The patent transitions from tabular or numerical data presentation to graphical visual representation, adding a visual dimension to data display. This dimensional change enables detailed prediction data to be understood more easily through visual patterns, trends, and comparisons rather than requiring complex data processing or analysis systems.
Data Source
AI summary
A prediction data display device includes a processor configured to input, into a trained model obtained by performing a training process on a model with training data including information on first characteristic data indicating a deterioration degree of a first consumable in a first period as input data and information on second characteristic data indicating a deterioration degree of the first consumable in a second period after the first period as ground-truth data, information on third characteristic data indicating a deterioration degree of a second consumable in the first period to calculate information on fourth characteristic data indicating a deterioration degree of the second consumable in the second period; and display a graph of the fourth characteristic data and a graph of the second characteristic data with a display appearance determined in accordance with a similarity between the information on the first characteristic data and the third characteristic data.