Stress Estimation Model Using Chronobiological Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current stress estimation models face challenges in achieving high accuracy due to the lack of appropriate feature quantities for machine learning, which affects the precision of stress level estimation.
Innovation Solution
An information processing apparatus and method that identifies a time zone of interest with notable chronic stress tendencies in biological signals and extracts relevant feature quantities from these signals for use in machine learning models to enhance stress level estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If feature quantities are extracted from all biological signals over a predetermined time period, then the quantity of training data is increased, but the accuracy of stress estimation is reduced due to inclusion of irrelevant data
Solution Approach 1:
The patent segments the predetermined time period into multiple time zones based on chronobiological characteristics (e.g., circadian rhythm phases). Feature quantities are extracted separately from each time zone rather than treating all data uniformly. This segmentation allows the system to focus on biologically relevant periods while excluding irrelevant data, thereby improving estimation accuracy without sacrificing data quantity.
Solution Approach 2:
The patent applies local quality by assigning different extraction conditions to different time zones. Each time zone is treated with specific extraction parameters tailored to its chronobiological characteristics. This ensures that feature quantities are extracted with appropriate quality standards for each temporal region, improving overall estimation accuracy while maintaining comprehensive data utilization.
2Productivity
If feature quantities are extracted from the entire time period, then data utilization is maximized, but the relevance of extracted features to chronic stress is reduced
Solution Approach 1:
The patent performs preliminary action by pre-defining time zones based on chronobiological knowledge before extracting feature quantities. The time zones are predetermined according to circadian rhythm patterns and other biological cycles. This preliminary structuring ensures that only data from chronobiologically relevant periods are processed, improving the reliability and relevance of extracted features while maintaining efficient data utilization.
Solution Approach 2:
The patent changes parameters by adjusting extraction conditions according to different time zones. Extraction parameters such as feature types, processing methods, and thresholds are modified based on the chronobiological characteristics of each time zone. This parameter adaptation ensures that feature extraction is optimized for each temporal region, enhancing both relevance to chronic stress and data utilization efficiency.
Data Source
AI summary
In order to appropriately extract feature quantities for use in machine learning or estimation of a stress level, an information processing apparatus (1, 4) includes: an identification means (11, 404) of identifying, as a time zone of interest, a time zone in which a chronic stress tendency is notably shown in biological signals which have been acquired from a subject over a predetermined time period; and an extraction means (12, 405) of extracting one or more feature quantities from biological signals acquired in the time zone of interest which has been identified, the one or more feature quantities being used in machine learning of an estimation model for estimating a stress level or used in estimation of a stress level using the estimation model.