Model Update Determination Using Data Difference Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional model update determination methods are inefficient and inaccurate, as they fail to consider the combination of overall, structural, and confidence differences between historical and new data items, leading to suboptimal model maintenance.
Innovation Solution
A method that determines overall, structural, and confidence differences between historical and new data items to provide an indication for model updates, using clustering techniques and confidence intervals to assess the significance of these differences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional model update determination methods are used, then the process is simple, but the accuracy and efficiency of model update determination deteriorates
Solution Approach 1:
The determination method is segmented into three distinct difference calculations: overall difference (comparing data distributions), structural difference (comparing feature relationships), and confidence difference (comparing prediction uncertainties). Each segmentation addresses a specific aspect of model update determination, improving comprehensive accuracy while maintaining modular complexity management.
Solution Approach 2:
The patent introduces multiple dimensional assessments beyond traditional single-metric determination. By evaluating overall difference, structural difference, and confidence difference as separate dimensions, the method captures nuanced changes in data and model behavior, significantly improving determination accuracy through multi-dimensional analysis.
2Productivity
If multiple difference factors are considered, then the determination accuracy improves, but the computational complexity increases
Solution Approach 1:
The patent performs preliminary clustering of historical data and new data into comparable groups before calculating differences. This preliminary organization enables more efficient computation of overall, structural, and confidence differences by reducing the search space and enabling vectorized operations, thereby improving productivity while managing computational complexity.
Solution Approach 2:
The method uses confidence intervals as copied statistical representations of data distributions and prediction uncertainties. By working with these statistical copies rather than raw data, the system efficiently computes confidence differences without processing entire datasets, improving computational efficiency while maintaining determination accuracy.
Data Source
AI summary
According to embodiments, a method, a device and a computer program product for model update determination are proposed. In the method, a plurality of historical data items and a plurality of new data items are obtained. The plurality of historical data items were used for training a model, and the plurality of new data items are to be applied to the model. At least one of an overall difference, a structural difference, and a confidence difference between the plurality of historical data items and the plurality of new data items is determined. Thereby, an indication indicating whether to update the model is determined based on the at least one of the overall difference, the structural difference, and the confidence difference.