Automated Variable Selection for Machine Learning Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional machine-learning models require a large number of variables, leading to increased computational resources and reduced prediction accuracy due to the inclusion of non-predictive variables, making it difficult for human users to manually select relevant variables, especially with large datasets.
Innovation Solution
An automated machine-learning optimization routine iteratively removes variables from the model, comparing outputs to determine which variables have a minimal impact below a predetermined threshold, resulting in an optimized model that operates with a reduced set of necessary variables for efficient predictive analyses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a large number of variables are included in the machine-learning model, then the model can potentially capture more patterns from the data, but the computational resources required increase and prediction accuracy decreases due to non-predictive variables
Solution Approach 1:
The patent extracts and removes non-predictive variables from the machine-learning model through automated variable selection. The system evaluates each variable's contribution to prediction accuracy and eliminates those that do not meet a predetermined threshold, thereby reducing computational resources while maintaining or improving prediction accuracy.
Solution Approach 2:
The patent changes the parameter of variable count in the model by dynamically selecting which variables to include based on their predictive value. The system adjusts the model's variable composition iteratively, removing variables that contribute minimally to prediction accuracy, thus optimizing the balance between computational efficiency and predictive performance.
2Reliability
If a large number of variables are included in the machine-learning model, then the model may capture more data patterns, but the device complexity increases making it difficult for human users to manually select relevant variables
Solution Approach 1:
The patent implements self-service by enabling the machine-learning model to automatically select and evaluate its own variables without human intervention. The system autonomously determines which variables are predictive and which are not, eliminating the need for manual variable selection and reducing the complexity burden on human users.
Solution Approach 2:
The patent employs feedback mechanisms where the model's prediction accuracy is continuously evaluated as variables are added or removed. This feedback loop allows the system to automatically adjust the variable set, identifying the optimal combination that maximizes prediction accuracy while minimizing model complexity.
3Productivity
If variables are iteratively removed from the model to reduce computational resources, then the model efficiency improves, but the time required for model optimization increases
Solution Approach 1:
The patent applies preliminary action by pre-evaluating variables and establishing criteria for their inclusion or exclusion before final model deployment. The system performs initial variable screening and removes obviously non-predictive variables upfront, reducing the iterative optimization process and minimizing the time required to achieve an efficient model.
Data Source
AI summary
Computer-implemented systems and methods for reducing an amount of computational resources consumed by a machine-learning model are provided. A machine-learning model is applied to a dataset to generate a first output. The machine-learning model includes a plurality of variables. Variables are iteratively removed from the machine-learning model, and for each iteration, the machine-learning model is applied with one or more variables removed from the dataset to generate a second output. For each iteration, the first and second outputs are compared. A subset of the removed variables having impact below a predetermined threshold on an output of the machine-learning model is determined based on the comparisons. An optimized machine-learning model that omits the subset of variables is applied to new data to generate an output for the new data.