Reversal Point Detection in Time Series Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In large online networks, identifying and ranking significant reversal points in time series data is challenging due to the vast amount of data and the computational expense of existing methods, which do not scale effectively and fail to analyze the importance of reversal points.
Innovation Solution
A machine-learning model is trained to detect and rank reversal points in discrete time series using a first-order derivative, evaluating them based on abnormality and significance, with a scoring algorithm that considers the length of the previous trend, number of reversal points, and magnitude changes, allowing for scalable analysis in large online networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to analyze time series data in large online networks, then comprehensive data analysis is achieved, but computational complexity and processing time become prohibitively high
Solution Approach 1:
The patent extracts and focuses only on reversal points (extremum points) in time series data rather than analyzing all data points. By identifying and isolating these critical points where the first-order derivative changes sign, the system reduces the vast amount of data to be processed to only the most significant events, thereby reducing computational complexity while maintaining detection accuracy.
Solution Approach 2:
The patent segments the time series data analysis into distinct phases: identifying reversal points through first-order derivative calculations, evaluating abnormality metrics for each reversal point, and ranking them by significance. This segmentation allows each component to be processed independently and efficiently, reducing overall computational burden.
2Loss of information
If all potential metrics and values are analyzed in large online networks, then complete information is obtained, but the difficulty of determining important metrics increases and processing becomes inefficient
Solution Approach 1:
The patent changes the parameter of analysis from examining all metric values to examining only reversal points where the first-order derivative changes sign. This parameter transformation converts a comprehensive but inefficient analysis into a focused analysis of critical events, maintaining information completeness about significant changes while dramatically improving processing efficiency through the abnormality evaluation and ranking system.
3Reliability
If existing methods are applied to large datasets, then data analysis coverage is maintained, but scalability to large online networks is lost
Solution Approach 1:
The patent replaces traditional mechanical data processing approaches with a mathematically efficient system based on first-order derivative calculations and abnormality metrics. This substitution enables the system to scale to large online networks while maintaining analysis reliability, as the mathematical operations required are computationally efficient and can be applied consistently across datasets of any size.
Data Source
AI summary
In an example embodiment, a model is trained to specifically identify reversal points in data and then to rank these reversal points in order of importance. A reversal point shall be defined as a point in which a particular metric, specifically a first order derivative, crosses over from positive to negative or vice-versa. Users are more likely to be interested in abnormal and significant changes in data, and thus the machine-learned model is trained to evaluate a reversal point based on two dimensions: abnormality and significance.