International business machines corporation (20240160773). DATA MINIMIZATION USING GLOBAL MODEL EXPLAINABILITY simplified abstract
Contents
DATA MINIMIZATION USING GLOBAL MODEL EXPLAINABILITY
Organization Name
international business machines corporation
Inventor(s)
Abigail Goldsteen of Haifa (IL)
DATA MINIMIZATION USING GLOBAL MODEL EXPLAINABILITY - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240160773 titled 'DATA MINIMIZATION USING GLOBAL MODEL EXPLAINABILITY
Simplified Explanation
An embodiment of the invention involves analyzing a predictive model and its input data using an explainability algorithm to determine the importance of a feature. The embodiment then generalizes the feature values using a generalization function and identifies an alternative feature based on a set of candidate feature values. If the accuracy of the predictive model meets a threshold performance value, the embodiment maps the feature values in the input data to a generalized representative value in the generalized domain.
- Explanation of the patent/innovation:
- Analyzing predictive models and input data using explainability algorithms - Generalizing feature values with a generalization function - Identifying alternative features based on candidate feature values - Mapping feature values in input data to a generalized representative value
Potential applications of this technology: - Improving the interpretability of predictive models - Enhancing feature selection processes in machine learning algorithms
Problems solved by this technology: - Addressing the lack of transparency in predictive models - Streamlining the feature selection process in machine learning
Benefits of this technology: - Increased understanding of model predictions - Improved model performance through feature generalization
Potential commercial applications of this technology: - Predictive analytics software - Machine learning platforms
Possible prior art: - Previous methods for feature selection in machine learning algorithms
Questions: 1. How does this technology compare to existing feature selection techniques in machine learning? 2. What impact does feature generalization have on the overall performance of predictive models?
Original Abstract Submitted
an embodiment analyzes a predictive model and its input data for the predictive model using an explainability algorithm resulting in a feature importance value of a feature. the embodiment analyzes feature values of the feature using a generalization function resulting in a set of candidate feature values. the embodiment determines an alternative feature based on the set of candidate feature values, wherein the alternative feature is a generalization of the feature. the embodiment compares an accuracy of the predictive model to a threshold performance value and, responsive to the accuracy being above the threshold performance value, maps feature values in the input data that are in the set of candidate feature values to a generalized representative value in the generalized domain.
