Synthetic Data Generation via Static-Dynamic Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for generating synthetic data for predictive analytics often require large amounts of processing power and may not retain trend and seasonal information, leading to inaccurate predictions due to limited or sensitive data availability.
Innovation Solution
A decompose-merge model is used to break down a base dataset into static and dynamic components, allowing for the generation of synthetic data that maintains time-dependent characteristics, such as trends and seasonality, by replacing the dynamic component with a new synthetic one while preserving the static components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large amount of actual analytics data is collected to produce accurate predictions, then prediction accuracy is improved, but processing power and processing time requirements increase significantly
Solution Approach 1:
The patent creates synthetic copies of actual analytics data by decomposing real data into static and dynamic components, then regenerating synthetic datasets that replicate the statistical properties and patterns of the original data without requiring access to or processing of the full original dataset. This copying approach enables accurate predictions using synthesized data pools.
Solution Approach 2:
The patent segments the analytics data into distinct static components (time-independent characteristics) and dynamic components (time-dependent variations including trends and seasonality). This segmentation allows selective replication of only the essential pattern-carrying elements rather than copying the entire dataset, reducing processing requirements while maintaining prediction accuracy.
2Quantity of substance
If conventional autoregressive models are used to generate synthetic data, then data pool expansion is achieved, but time dependent characteristics such as trends and seasonality are lost
Solution Approach 1:
The patent segments the data into static components (mean, variance) and dynamic components (trends, seasonality, random variations). By separately modeling and preserving the dynamic components during synthetic data generation, the method expands the data pool while retaining crucial time-dependent characteristics that conventional autoregressive models discard.
Solution Approach 2:
The patent transforms the data representation by changing parameters from raw time series values to decomposed components (trend, seasonal, random). This parameter transformation enables the synthetic data generation process to explicitly preserve time-dependent characteristics while still achieving data pool expansion through stochastic replication of the component structures.
3Reliability
If the available analytics data pool is limited or contains sensitive information, then data privacy is protected, but prediction accuracy deteriorates due to insufficient data points
Solution Approach 1:
The patent creates synthetic copies of the limited or sensitive analytics data by replicating its statistical properties and temporal patterns. These synthetic copies expand the effective data pool for prediction modeling without exposing or requiring access to the original sensitive data, thus maintaining privacy protection while improving prediction accuracy through increased data availability.
Data Source
AI summary
Methods for generating synthetic data based on time dependent data with increased accuracy include decomposing a base dataset into a base dynamic component and at least one static component. Decomposing the base dataset includes applying a decomposition model to the base dataset. One or more embodiments generate a synthetic dynamic component based on the base dynamic component. One or more embodiments merge the synthetic dynamic component with the at least one static component to generate a synthetic dataset having at least some of the time dependent characteristics of the base dataset.