Variable Importance Ranking in Nonparametric Machine Learning Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-parametric machine learning models, which can utilize an infinite number of independent variables, lack transparency in identifying the most significant variables contributing to outcomes, making it difficult for businesses and researchers to focus their resources effectively.
Innovation Solution
A system and method that processes raw values through a statistical model to determine probability values, cumulative distributions, and rank variables, allowing for the identification of the relative importance of variables by binning them into equally sized buckets and assigning rank numbers based on mean probability values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If non-parametric machine learning models are used to utilize an infinite number of independent variables, then the model can handle more data and model more complicated relationships, but the transparency in identifying significant variables is lost
Solution Approach 1:
The patent introduces an intermediary system that processes the output of non-parametric machine learning models to extract and rank variable significance. This intermediary layer translates the complex internal workings of models like gradient boosting machines into interpretable importance metrics, allowing users to understand which variables drive predictions without sacrificing the model's ability to handle infinite variables.
Solution Approach 2:
The patent replaces the need for direct human interpretation of complex model internals with an automated computational system. Instead of manually analyzing model behavior, the system automatically calculates variable importance through structured processing of model outputs, substituting mechanical analysis with algorithmic evaluation.
2Reliability
If more independent variables are included in the model to improve accuracy, then more data can be leveraged, but processing time and computational resources increase
Solution Approach 1:
The patent extracts only the essential information needed for variable importance ranking from the full model output. By taking out and processing only the relevant probability values and predictions rather than analyzing the entire model computation, the system achieves variable significance identification without proportionally increasing processing time.
Solution Approach 2:
The patent performs partial action by calculating variable importance based on a subset of model outputs rather than re-running the entire model. The system uses the predictions already generated by the model and processes only the necessary components to derive importance metrics, avoiding excessive computation while still providing comprehensive variable analysis.
Data Source
AI summary
Systems and methods for determining relative importance of one or more variables in a non-parametric model include: receiving, raw values of the variables corresponding to one or more entities; processing the raw values using a statistical model to obtain probability values for the variables and an overall prediction value for each entity; determining a plurality of cumulative distributions for the variables based on the raw values and the number of entities having a specific raw value; grouping the variables into a plurality of equally sized buckets based on the cumulative distributions; determining a mean probability value for each bucket; assigning a rank number for each bucket based on the mean probability values; compiling a table for the entities based on the raw values and the buckets corresponding to the raw values; and determining the relative importance of the variables for the entities based on the rank numbers.