Time Series Correlation Detection Platform
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data analysis methods for organizations are inefficient and prone to human subjectivity, taking weeks or months to detect correlations across various aspects of an organization, leading to out-of-date and inaccurate results, especially when dealing with large and complex data sets.
Innovation Solution
A correlation detection platform utilizing cloud computing and machine learning techniques to automate the process of detecting correlations in time series data, capable of processing millions or billions of data elements, by validating, classifying, generating charts, and applying correlation detection methods to identify correlations quickly and objectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional manual data analysis methods are used, then human expertise and judgment can be applied, but the process takes weeks or months and is prone to subjectivity
Solution Approach 1:
The patent replaces manual mechanical analysis processes with automated computational systems. Machine learning models and algorithms automatically detect correlations in time series data, eliminating human subjectivity and dramatically reducing analysis time from weeks/months to minutes/hours while maintaining or improving detection accuracy.
Solution Approach 2:
The system enables self-service correlation detection where the automated platform independently processes data, applies multiple correlation detection techniques, and generates results without requiring continuous human intervention. The system serves itself by automatically selecting and applying appropriate analysis methods based on the data characteristics.
2Quantity of substance
If traditional analysis methods are used, then computing resources are conserved, but the analysis cannot handle large and complex data sets efficiently
Solution Approach 1:
The patent segments the complex correlation detection task into multiple independent components: data preprocessing, multiple correlation detection techniques (Pearson, Spearman, Kendall tau), and result aggregation. This segmentation allows the system to handle large datasets by processing them through specialized modules, each optimized for specific aspects of correlation analysis.
Solution Approach 2:
The platform is designed with multi-functionality to handle various types of time series data and apply multiple correlation detection techniques universally. The system can process different data formats, scales, and patterns using the same core infrastructure, enabling it to manage large and diverse datasets without proportionally increasing complexity.
3Reliability
If comprehensive correlation detection is performed on all data pairs, then complete analysis is achieved, but computing resources are excessively consumed
Solution Approach 1:
The patent applies partial action by selectively applying different correlation detection techniques to different data pairs based on their characteristics. Instead of exhaustively applying all techniques to all possible pairs, the system intelligently selects appropriate methods for each pair, achieving sufficient analysis coverage while significantly reducing computing resource consumption.
Solution Approach 2:
The system dynamically changes parameters such as the selection of correlation detection techniques, data sampling rates, and analysis thresholds based on the specific characteristics of each dataset. This adaptive parameter adjustment allows the platform to maintain reliable correlation detection while optimizing computing resource usage for each specific analysis task.
Data Source
AI summary
A device may receive time series data from one or more data sources. The device may pre-process the time series data to generate a respective time series chart for multiple classifications included in the time series data. The device may randomly select a pair of respective time series charts and a correlation detection technique after pre-processing the time series data. The device may determine a correlation for the pair of respective time series charts based on using the correlation detection technique. The device may determine a score for the pair of respective time series charts based on the correlation of the pair of respective time series charts. The score may indicate an extent to which the pair of respective time series charts is correlated with each other. The device may perform one or more actions after determining the score for the pair of respective time series charts.