Matrix Decomposition for Anomaly Detection in Measurement Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Analyzing measurement results from complex target systems, such as communications networks or industrial processes, is challenging due to the vast amount of data, making it difficult to identify the most relevant anomalous results effectively.
Innovation Solution
A computer-implemented method involving matrix decomposition to separate normal and anomalous measurement results, where a trained model decomposes initial measurement data into stable and unstable matrices, and subsequent comparison with new data to identify anomalies by subtracting matching subsets, enabling efficient evaluation of system performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional anomaly detection models are used to analyze measurement results, then anomalies can be identified, but the analysis becomes slow and computationally intensive due to the vast amount of data
Solution Approach 1:
The patent segments the measurement data into two separate matrices: a first matrix containing normal measurement results and a second matrix containing anomalous measurement results. This segmentation allows the system to process and compare only relevant portions of data, significantly reducing computational complexity and analysis time while maintaining anomaly detection accuracy.
Solution Approach 2:
The patent performs preliminary decomposition of the measurement data into normal and anomalous components before comparison. By pre-processing the data to separate normal patterns from anomalies, the system eliminates the need for computationally intensive real-time analysis of entire datasets, thus reducing analysis time while preserving detection precision.
2Reliability
If comprehensive measurement data is collected for accurate analysis, then system performance evaluation is thorough, but the complexity of data processing increases
Solution Approach 1:
The patent divides comprehensive measurement data into structured matrices that separate normal operational patterns from anomalous events. This segmentation simplifies the processing complexity by allowing independent analysis of each matrix type, while maintaining reliable system performance evaluation through comprehensive coverage of both normal and abnormal conditions.
Solution Approach 2:
The patent transforms raw measurement data into a different parameter representation through matrix decomposition, converting complex multivariate data into structured matrices with distinct normal and anomalous components. This parameter transformation reduces processing complexity while preserving the reliability needed for accurate system performance evaluation.
Data Source
AI summary
Analyzing measurement results of a target system. The analysis is performed by receiving a first matrix including first measurement results of the target system; training a matrix decomposition model with the first matrix to obtain a third matrix of normal or stable measurement results and a fourth matrix of anomalous or unstable measurement results; receiving a second matrix including second measurement results of the target system, wherein the second measurement results are later measurement results compared to the first measurement results; selecting from the third matrix a subset that matches with the second matrix; subtracting the selected subset from the second matrix to obtain a fifth matrix; outputting the fifth matrix or information derived from the fifth matrix for the purpose of evaluating performance of the target system.