Sequential Data Gap Handling via Sliding Window Pattern Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for handling data gaps in sequential data often neglect or inadequately utilize available information, failing to capture interactions among variables and relying on assumptions that limit their effectiveness, especially in scenarios where data gaps occur across multiple variables or over time.
Innovation Solution
A method that utilizes a sliding window based on timestamps to extract patterns using the Random Convolutional Kernel Transform (ROCKET) algorithm, masking data gaps and feeding modified data into a prediction model to fill gaps, thereby leveraging all available information without prior assumptions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods are used to handle data gaps, then the processing is simple, but the accuracy and effectiveness are limited due to neglecting available information and relying on assumptions
Solution Approach 1:
The patent segments the data processing into distinct components: identifying data gaps, determining sliding windows for each gap, extracting patterns from dependent data within those windows, and generating predictions. This segmentation allows complex pattern extraction to be applied only where needed (at data gaps) rather than to entire datasets, improving accuracy while managing complexity through modular processing
Solution Approach 2:
The patent introduces an intermediary pattern extraction step that transforms raw dependent data into extracted patterns, which then serve as input for the prediction model. This intermediary layer enables the system to capture variable interactions without requiring the prediction model to directly process raw data, thereby improving accuracy while maintaining manageable complexity through abstraction
2Measurement precision
If all available data is utilized to fill data gaps, then the accuracy improves, but the computational complexity increases
Solution Approach 1:
The patent applies local quality by focusing computational resources on local regions around data gaps rather than processing entire datasets uniformly. By determining sliding windows specific to each data gap and extracting patterns only from dependent data within those windows, the system achieves high accuracy at gaps while minimizing overall computational power requirements
Solution Approach 2:
The patent implements partial action by extracting patterns from only the necessary portion of data (dependent data within sliding windows) rather than processing all available data. This selective approach provides sufficient information for accurate gap filling without the excessive computational cost of analyzing entire datasets
3Measurement precision
If sliding windows with dependent data are used, then the variable interactions are captured, but the data processing complexity increases
Solution Approach 1:
The patent segments the data processing into distinct components: identifying data gaps, determining sliding windows for each gap, extracting patterns from dependent data within those windows, and generating predictions. This segmentation allows complex pattern extraction to be applied only where needed (at data gaps) rather than to entire datasets, improving accuracy while managing complexity through modular processing
Solution Approach 2:
The patent performs preliminary action by pre-determining sliding windows and extracting patterns from dependent data before generating final predictions. This preliminary pattern extraction prepares the data in advance, allowing the prediction step to focus solely on filling gaps using the pre-processed patterns, thereby capturing variable interactions while managing processing complexity through staged computation
Data Source
AI summary
A method, a computer program product, and a computer system handle a data gap in sequential data. The method includes receiving the sequential data for a period of time. The method includes selecting the data gap in the sequential data at a timestamp. The method includes determining a sliding window associated with the data gap based on the timestamp for a duration of time. The sliding window includes dependent data from which the data gap depends. The method includes, as a result of the dependent data of the sliding window including at least one window data gap, generating extracted patterns based on the dependent data to mask the at least one window data gap. The method includes determining a prediction to fill the data gap using a prediction model that takes as input modified data based on the dependent data and the extracted patterns.