Brainwave Training Data Augmentation via Covariance Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for objectively evaluating pain using brainwaves face challenges due to the subjective nature of pain intensity and variability in brainwave signals, lacking effective methodologies for monitoring temporal changes and efficient model generation for pain differentiation.
Innovation Solution
A system and method for augmenting supervisory data through covariance matrix decomposition and application of random numbers to enhance brainwave data for pain classification and differentiation, utilizing techniques like Cholesky, LU, and QR decompositions, and random number application to maintain data relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If more supervisory data are collected to improve machine learning accuracy, then prediction accuracy is improved, but the burden on organisms increases due to more stimulations required
Solution Approach 1:
The patent creates synthetic copies of real brainwave data by applying covariance matrix decomposition and random number generation. The system decomposes the covariance matrix of real data into Q*Q' form, then generates augmented data by applying random numbers to the decomposed matrix, creating realistic synthetic samples that replicate the statistical properties of actual brainwave responses without requiring additional stimulations of the organism.
Solution Approach 2:
The patent transforms the data representation by changing parameters through covariance matrix decomposition. By decomposing the covariance matrix into Q*Q' form and applying random numbers to the decomposed components, the system generates new data samples with preserved statistical characteristics. This parameter transformation allows synthetic data generation that maintains the essential relationships and patterns of real brainwave data while expanding the dataset without additional biological sampling.
2Measurement precision
If more brainwave data are collected to improve pain classification accuracy, then measurement precision is improved, but data collection complexity increases
Solution Approach 1:
The system generates synthetic brainwave data by copying the statistical structure of real data through covariance matrix decomposition. The decomposition into Q*Q' form allows the system to replicate the correlation patterns and variability of actual pain response data, creating augmented datasets that expand training samples without requiring additional complex data collection procedures or increased measurement complexity.
3Quantity of substance
If simple data augmentation methods are used to increase sample size, then quantity of data is improved, but data relationships may be compromised
Solution Approach 1:
The patent applies parameter changes through covariance matrix decomposition to ensure data relationship integrity. By decomposing the covariance matrix into Q*Q' form and applying random numbers to the decomposed components, the system preserves the statistical structure, correlation patterns, and relationships inherent in the original data. This mathematical approach ensures that augmented samples maintain the same distributional properties and relationships as the original data, preventing the introduction of artificial correlations or distortions.
Data Source
AI summary
The problem to be solved is to provide a system and the like for augmenting supervisory data while maintaining the relationship among a plurality supervisory data used for machine learning. The present disclosure provides a system for augmenting supervisory data used for machine learning, the system including an obtaining means that obtains a plurality of supervisory data, a first processing means that derives a covariance matrix from the plurality of supervisory data, a second processing means that decomposes the covariance matrix, and a third processing means that applies a random number to the decomposed matrix.