Reduced Input Space Construction via Rejected Variable Dimensionality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data modeling techniques face challenges in selecting the optimal set of variables for analysis while balancing interpretability and information retention, often resulting in the loss of significant data when rejecting variables.
Innovation Solution
A computer-implemented system that performs variable predictiveness determination on candidate variables, selects a subset based on these values, generates derived variables from rejected variables using dimension reduction techniques, and evaluates their predictiveness to include them in the data model, thereby enhancing the model's accuracy and interpretability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If variables are selected based on predictiveness to improve model accuracy, then prediction accuracy is improved, but the number of variables increases and interpretability deteriorates
Solution Approach 1:
The patent segments the variable selection process into multiple stages: first selecting variables based on predictiveness, then applying dimension reduction techniques to create derived variables from rejected variables. This segmentation allows the model to achieve high accuracy through comprehensive variable utilization while managing complexity through structured processing
Solution Approach 2:
The patent applies dimension reduction techniques (such as principal components analysis or factor analysis) to transform the rejected variables into a smaller number of derived variables. This dimensional transformation reduces the complexity of the variable set while preserving the information content, thereby resolving the contradiction between accuracy and interpretability
2Loss of information
If more variables are selected to retain significant data information, then information retention is improved, but model interpretability deteriorates
Solution Approach 1:
The patent extracts meaningful patterns and information from the rejected variables using dimension reduction techniques. By taking out the essential information from the rejected variable space and transforming it into derived variables, the system retains significant data information while presenting it in a more interpretable format
Solution Approach 2:
The patent changes the parameters of the variable representation by transforming original variables into derived variables through dimension reduction. This parameter transformation preserves information content while altering the variable structure to improve interpretability
3Device complexity
If dimension reduction techniques are applied to rejected variables to reduce complexity, then model complexity is reduced, but prediction accuracy may deteriorate
Solution Approach 1:
The patent discards the rejected variables in their original form but recovers their information content through dimension reduction techniques. By recovering the essential information from rejected variables and transforming it into derived variables, the system maintains prediction accuracy while reducing model complexity
Data Source
AI summary
Computer-implemented systems and methods are provided for generating a data model. A variable predictiveness determination is performed on the population of candidate variables. A plurality of variables from the population of candidate variables are selected as a selected set based on the variable predictiveness values. A plurality derived variables are generated based on variables in the rejected set without consideration of any variables in the selected set. One or more derived variables are selected as based on derived variable predictiveness values of the derived variables, and the selected set and the one or more selected derived variables are stored as the model input variables for the data model.