Dynamic Prediction Model Using Data Integration and Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Establishing a prediction model with high accuracy and low cost is challenging due to the high cost of obtaining target data and the incompleteness and inaccuracies of auxiliary data, leading to models with high bias and uncertainty.
Innovation Solution
A dynamic prediction model establishment method using a two-stage model stacking technology that integrates auxiliary data sets with target data sets, modifies the integration model based on error and uncertainty degrees, and provides sampling point recommendations to reduce error and uncertainty through a processing device and user interface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If target data is used to establish a prediction model, then prediction accuracy is improved, but acquisition cost increases significantly
Solution Approach 1:
The patent combines auxiliary data sets with target data sets to establish an integration model. By merging multiple data sources (auxiliary data from simulators and historical records with actual target data), the model achieves high prediction accuracy while reducing dependency on expensive target data alone. The integration model leverages the strengths of both data types to improve overall model performance.
Solution Approach 2:
The patent introduces an integration model as an intermediary between auxiliary data and final predictions. This integration model acts as a mediator that processes both auxiliary and target data, modifying the auxiliary-based predictions using target data to reduce bias and uncertainty. The integration model serves as a bridge that combines information from multiple sources to achieve accurate predictions with reduced cost.
2Quantity of substance
If auxiliary data is used to establish a prediction model, then acquisition cost is reduced, but prediction accuracy deteriorates due to simulator incompleteness and product differences
Solution Approach 1:
The patent implements a feedback mechanism where the integration model is modified using target data to correct errors in auxiliary data-based predictions. The system continuously refines the model by incorporating feedback from actual target data, reducing the bias and uncertainty inherent in auxiliary data alone. This feedback loop enables the model to learn from discrepancies between auxiliary and target data, improving accuracy over time.
Solution Approach 2:
The patent modifies the integration model by changing parameters based on target data to reduce error degrees and uncertainty degrees. The system adjusts model parameters dynamically, transforming the static integration model into a dynamic prediction model that adapts to reduce the gaps between auxiliary data predictions and actual target outcomes. This parameter adjustment process directly addresses the accuracy limitations of auxiliary data.
3Reliability
If a large volume of target data is obtained to reduce model uncertainty, then prediction accuracy is improved, but data acquisition cost and time increase
Solution Approach 1:
The patent applies partial action by using a relatively small amount of target data to modify the integration model, rather than requiring large volumes of target data. The integration model, pre-trained on auxiliary data, provides a strong baseline that requires minimal target data refinement. This approach achieves significant uncertainty reduction with limited target data collection, saving time and resources.
Solution Approach 2:
The patent performs preliminary action by pre-establishing the integration model using auxiliary data before final prediction deployment. The model is pre-trained and refined in advance using available auxiliary data sets, so that when actual target data becomes available, the model can be quickly adjusted with minimal data collection time. This preliminary preparation reduces the need for extensive real-time data acquisition.
4Quantity of substance
If simulator-based auxiliary data is used, then data acquisition cost is reduced, but data completeness and accuracy deteriorate due to simulation limitations
Solution Approach 1:
The patent creates a composite data structure by combining auxiliary data from multiple sources (simulators, historical records) with target data. This composite approach compensates for the incompleteness of individual auxiliary data sources. By integrating diverse data types and sources, the model achieves more complete and accurate representations than any single auxiliary source could provide alone.
Data Source
AI summary
A dynamic prediction model establishment method, an electric device and a user interface are provided. The dynamic prediction model establishment method includes the following steps. An integration model is established by a processing device according to at least one auxiliary data set. The integration model is modified as a dynamic prediction model by the processing device according to a target data set. A sampling point recommendation information is provided by the processing device according to an error degree or an uncertainty degree between the at least one auxiliary data set and the target data set.