Time-Series Anomaly Detection Using Regularity Checks and Adaptive Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems struggle with detecting anomalies in time-series data due to the complexity and resource intensity of identifying seasonality patterns, leading to overwhelming alerts, false positives, and false negatives, especially when dealing with vast amounts of raw data.
Innovation Solution
Anomaly detection system that analyzes time-series data to determine regularity and seasonality patterns, using silhouette scores to select appropriate thresholds and generate anomaly bands, with adaptive methods to handle varying data sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex seasonality pattern identification methods are used to accurately detect anomalies, then measurement precision is improved, but device complexity and computational resources increase significantly
Solution Approach 1:
The patent segments the anomaly detection process into distinct phases: data regularity assessment, seasonality pattern identification (using multiple candidate patterns), and anomaly detection. This segmentation allows the system to handle complexity in a structured manner, processing different aspects of the data separately rather than attempting a monolithic analysis.
Solution Approach 2:
The system dynamically adapts its detection approach based on the assessed regularity of the time-series data. When data exhibits high regularity, more sophisticated seasonality-based methods are applied; when regularity is low, simpler methods are used. This dynamic adaptation optimizes computational resources while maintaining detection accuracy.
2Reliability
If comprehensive analysis of all time-series data is performed to reduce false positives and false negatives, then reliability is improved, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary assessment of data regularity before committing to full anomaly detection processing. This preliminary action allows the system to identify suitable seasonality patterns in advance and determine the appropriate detection strategy, reducing the computational burden during the actual anomaly detection phase while maintaining reliability.
Solution Approach 2:
The system changes detection parameters based on the characteristics of the time-series data being analyzed. Different seasonality patterns (daily, weekly, monthly) are selected based on the data's temporal characteristics, allowing the system to optimize processing time while maintaining detection reliability for different types of data.
3Adaptability or versatility
If multiple seasonality patterns are evaluated to handle varying data sets, then adaptability is improved, but device complexity increases
Solution Approach 1:
The patent implements a universal anomaly detection framework that can handle multiple types of seasonality patterns (daily, weekly, monthly, and custom patterns) within a single system. The same core detection machinery adapts to different pattern types through configuration rather than requiring separate specialized systems, improving versatility without proportionally increasing complexity.
4Measurement precision
If adaptive thresholding methods are used to improve anomaly detection precision, then measurement precision is improved, but computational resources and processing complexity increase
Solution Approach 1:
The patent applies adaptive thresholding locally based on the specific characteristics of each time-series data set and its identified seasonality pattern. Rather than using a single global thresholding approach, the system tailors threshold parameters to match the local properties of the data, improving precision while avoiding the computational overhead of overly complex global adaptive methods.
Data Source
AI summary
Computerized methodologies are disclosed that are directed to detecting anomalies within a time-series data set. A first aspect of the anomaly detection process includes analyzing the regularity of the data points of the time-series data set and determining whether a data aggregation process is to be performed based on the regularity of the data points, which results in a time-series data set having data points occurring at regular intervals. A seasonality pattern may be determined for the time-series data set, where a silhouette score is computed to measure the quality of the fit of the seasonality pattern to the time-series data. The silhouette score may be compared to a threshold and based on the comparison, the seasonality pattern or a set of heuristics may be utilized in an anomaly detection process. When the seasonality pattern is utilized, the seasonality pattern may be utilized to generate thresholds indicating anomalous behavior.