Correlation Analysis for Accurate Abnormality Cause Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Invariant relation analysis struggles to accurately determine the abnormality cause when only a small number of correlations associated with the abnormality cause metric are destroyed, leading to erroneous identification of the metric causing the issue.
Innovation Solution
A system analysis device and method that calculates detection sensitivities for each correlation, using these sensitivities to identify the likelihood of correlation destruction and accurately extract the abnormality cause metric, rather than relying solely on the ratio of correlation destruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the ratio of correlation destruction is used to determine abnormality cause, then the determination process is simple, but the accuracy deteriorates when only a small number of correlations are destroyed
Solution Approach 1:
The patent changes the parameter used for determination from a simple ratio of correlation destruction to a composite index that incorporates detection sensitivity. This allows the system to maintain operational simplicity while significantly improving accuracy by weighting correlations based on their individual detection sensitivities rather than treating all correlations equally.
Solution Approach 2:
The patent introduces detection sensitivity as an intermediary parameter that mediates between the simple correlation destruction ratio and the final abnormality cause determination. This intermediary allows the system to refine the raw correlation data into a more accurate diagnostic indicator without complicating the overall determination process.
2Device complexity
If all correlations are analyzed equally to identify abnormality cause, then the method is straightforward, but erroneous identification occurs when correlation destruction is limited
Solution Approach 1:
The patent applies local quality by assigning different weights (detection sensitivities) to different correlations based on their individual characteristics. Instead of treating all correlations uniformly, the system identifies and emphasizes the locally significant correlations that have higher detection sensitivities, thereby improving reliability without requiring complex analysis of all correlations equally.
Solution Approach 2:
The patent transforms the uniform correlation analysis approach into a weighted analysis by changing the parameter from equal treatment of all correlations to differential weighting based on detection sensitivity. This parameter change enables the system to focus on the most informative correlations while maintaining reasonable methodological simplicity.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
In invariant relation analysis, an abnormality cause is accurately determined. A system analysis device (100) includes a correlation model storage unit (112) and an abnormality cause extraction unit (104). The correlation model storage unit (112) stores a correlation model (122) indicating a correlation of a pair of metrics in a system. The abnormality cause extraction unit (104) extracts a metric of candidate for an abnormality cause on the basis of a detection sensitivity calculated for each metric associated with a correlation for which correlation destruction is detected in correlations included in the correlation model (122). The detection sensitivity indicates a likelihood of occurrence of correlation destruction in each correlation associated with the metric at the time of abnormality of the metric.