Feature Vector Merging Rule Generation for ML Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine learning models in the medical field face challenges in predicting patient outcomes due to the generation of numerous feature vectors with similar meanings, leading to reduced prediction accuracy, especially when variations in notation and large unit sizes are considered, and manual merging of these vectors is time-consuming and unreliable.
Innovation Solution
A device and method for generating a data merging rule that specifies combinations of feature vectors with similar meanings by analyzing frequency distributions and similarity thresholds, allowing for automatic reduction of feature vector dimensions, thereby improving prediction accuracy without manual intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If feature vectors are generated by focusing only on difference in notation, then the number of feature vectors is increased, but the prediction accuracy is reduced due to generation of large number of feature vectors having the same or similar meaning
Solution Approach 1:
The patent merges feature vectors that have substantially the same or similar meaning into a single representative feature vector. This is achieved by identifying combinations of feature vectors with similar semantics and replacing them with a consolidated vector, thereby reducing the total number of feature vectors while maintaining or improving prediction accuracy through the use of a machine learning model that learns from the merged representations.
2Device complexity
If feature vectors are merged by manual operation of a person, then the number of dimensions of the feature vectors is reduced, but the time and effort required is significant and there is no guarantee that improvement in prediction accuracy can always be expected
Solution Approach 1:
The patent implements an automated system where the machine learning model itself performs the merging of feature vectors without requiring manual intervention. The model automatically identifies combinations of feature vectors with similar meanings and merges them, eliminating the need for manual operations and significantly reducing the time and effort required while providing consistent and reliable merging results based on data-driven criteria.
3Measurement precision
If the unit size of grouping for ages is made large, then the prediction accuracy is improved by grouping patients, but the prediction accuracy is lowered when the group is formed with an excessively large unit size
Solution Approach 1:
The patent dynamically determines the appropriate grouping size for feature vectors based on the actual data characteristics and model performance. Instead of using fixed or manually设定的 group sizes, the system adapts the grouping granularity automatically, allowing it to optimize the balance between grouping benefits and the loss of detail when group sizes become excessively large, thereby maintaining high prediction accuracy across different data scenarios.
Data Source
AI summary
There is provided a device for generating a data merging rule for a machine learning model, the device including: a processor, and a memory connected to or built in the processor, in which the processor is configured to execute specifying processing of specifying a combination of feature vectors that are included in a data set including a correct answer label and are allowed to be merged, and rule generation processing of generating a merging rule of the feature vectors based on a combination of the feature vectors that are allowed to be merged.