Error-Bounded Time Series Joins Via Compact Dictionaries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for similarity joins in time series data mining are inefficient and lack effective ways to improve computation speed and accuracy, particularly in inter-similarity joins.
Innovation Solution
A method involving a server computer that determines time series subsequences, computes initial and processed matrix profiles, forms a dictionary of subsequences, and uses this dictionary to generate an approximate matrix profile for anomaly detection, employing a compact dictionary representation to enhance computation efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional similarity join methods are used for time series data, then comprehensive analysis can be performed, but computation speed and efficiency deteriorate
Solution Approach 1:
The patent segments the time series data into subsequences and builds a dictionary of representative subsequences. Instead of comparing every subsequence against all others, the method segments the search space by using the dictionary to represent common patterns, thereby reducing the computational complexity while maintaining detection accuracy.
Solution Approach 2:
The patent introduces a dictionary of subsequences as an intermediary structure. This dictionary acts as a mediator between the query subsequence and the full time series database, enabling faster comparison by referencing pre-computed dictionary entries rather than performing exhaustive pairwise comparisons.
2Reliability
If exhaustive similarity join is performed on time series data, then all anomalies can be detected, but computational resources and time are excessively consumed
Solution Approach 1:
The patent performs preliminary actions by pre-building a dictionary of subsequences from the time series data before the actual anomaly detection process. This pre-processing step organizes the data into a searchable structure, eliminating the need for exhaustive comparisons during the detection phase and significantly reducing processing time.
Solution Approach 2:
The patent creates a compressed representation (copy) of the time series data in the form of a subsequence dictionary. This dictionary copy contains the essential patterns and structures of the original data, allowing anomaly detection to be performed on the compact representation rather than the full data set, thereby reducing computational overhead.
3Productivity
If approximate methods are used to improve computation speed, then processing efficiency increases, but detection accuracy may deteriorate
Solution Approach 1:
The patent replaces the mechanical exhaustive comparison system with a more efficient information-retrieval-based system. Instead of mechanically comparing every subsequence against all others, the method substitutes this with dictionary-based pattern matching and similarity searching, which leverages information theory and data structures to achieve faster results with maintained accuracy.
Data Source
AI summary
A method is disclosed. The method comprises determining a time series, a subsequence length. The length of the time series may then be determined, and an initial matrix profile may then be computed. The method may then form a processed matrix profile for a first subsequence of the subsequence length by applying the first subsequence to the initial matrix profile. A second subsequence may then be determined from the processed matrix profile. The method may then include comparing the second subsequence to other subsequences in a dictionary and adding it to the dictionary. The subsequences in the dictionary may be used to generate a plurality of subsequence matrix profiles. The method may then include forming an approximate matrix profile using the plurality of subsequence matrix profiles and then determining one or more anomalies in the time series or another time series using the approximate matrix profile.