Augmented Time Series for Adverse Event Timing Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current machine learning models face challenges in accurately predicting the timing of adverse events due to limitations in representing and processing time series data, which affects the adaptability of operating environments and the clarity of model explanations.
Innovation Solution
A system generates augmented time series data by augmenting historical panel data with missing observation values, increasing the frequency of observations, and using wavelet transforms to improve the granularity and accuracy of predictions, allowing for more effective timing predictions and adaptive operating environment modifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If historical panel data is used directly for training machine learning models, then the device complexity is reduced, but the measurement precision and prediction accuracy deteriorate due to missing observation values and low temporal granularity
Solution Approach 1:
The patent applies preliminary action by performing data augmentation before model training. Missing observation values are imputed using statistical methods (mean, median, mode imputation) and temporal interpolation to create complete time series datasets. This preprocessing step ensures that the training data is ready in advance, eliminating gaps that would otherwise reduce prediction accuracy without adding complexity during the modeling phase
Solution Approach 2:
The patent introduces an intermediary data transformation layer between raw historical panel data and the machine learning model. This intermediary process includes resampling techniques to standardize temporal frequencies, feature engineering to create derived variables, and data normalization. These intermediary steps bridge the gap between irregular historical data and the structured input requirements of prediction models, improving measurement precision while managing complexity through systematic transformation
2Measurement precision
If the frequency of observations in historical panel data is increased, then the prediction accuracy improves, but the loss of time and data processing requirements increase
Solution Approach 1:
The patent applies skipping by implementing efficient resampling strategies that selectively increase observation frequency only where necessary for accurate timing predictions. Rather than uniformly increasing frequency across all time points, the method identifies critical periods and events where higher granularity is needed, then applies targeted resampling. This approach rushes through less critical periods with lower frequency while maintaining high frequency during adverse events, reducing overall processing time while preserving prediction accuracy
Solution Approach 2:
The patent uses parameter changes by dynamically adjusting the temporal resolution of observations based on the specific prediction task and data characteristics. The system varies sampling frequencies across different time periods and variables, applying higher frequencies to critical predictors of adverse events and lower frequencies to stable variables. This parameter optimization balances measurement precision with processing efficiency, avoiding the time loss associated with uniformly high-frequency data collection
3Reliability
If data augmentation techniques are applied to impute missing values, then the reliability of predictions improves, but the device complexity increases due to additional processing steps
Solution Approach 1:
The patent applies self-service by implementing imputation methods that use the data's own internal structure and patterns to fill missing values. Statistical measures (mean, median, mode) are calculated from available observations within each variable, and temporal interpolation leverages the time-series nature of the data to predict missing points based on surrounding values. These self-service approaches improve prediction reliability by utilizing the data's inherent information without requiring external complex models or additional data sources, thereby limiting the increase in processing complexity
Data Source
AI summary
A system receives, from a remote computing device, a query for a timing of an adverse event associated with a target entity. The system determines, using a timing prediction model trained using a training process, the timing of the adverse event for the target entity from predictor variables associated with the target entity. The training process includes accessing an observational journal comprising historical panel data of the target entity including values of predictor variables for one or more time points and generating, from historical panel data, an augmented time series by augmenting the historical panel data with values of predictor variables for at time points for which the historical panel data does not include values of predictor variables. The system transmits to the remote computing device, a responsive message including at least the timing of the adverse event for use in controlling access of the target entity to one or more interactive computing environments.