| CPC G06F 40/40 (2020.01) [G06N 3/0455 (2023.01)] | 20 Claims |
|
1. A method for generating a contextually adaptable machine learning (ML) based classifier model, the method comprising:
obtaining a dataset including datapoints, feature values characterizing the datapoints according to input features, and labels classifying the datapoints according to a target label;
transforming each respective datapoint into a natural language statement (NLS), each NLS associating the respective datapoint's feature values with feature identifiers assigned to the corresponding input features, and each NLS associating the respective datapoint's label with a label identifier assigned to the target label;
generating a feature matrix for each NLS based on the feature identifiers and feature values;
transforming the feature matrix into a global feature vector;
generating a target vector for each NLS based on the label identifier and the corresponding label;
transforming the target vector into a global target vector having a same shape as the global feature vector; and
generating, using the global feature vector and the global target vector in conjunction with a similarity measurement operation and a loss function, an ML-based classifier model trained to generate, using one or more neural network models, a compatibility score predictive of an accuracy at which the classifier model can classify given data based on at least one of a different feature characterizing the given data or a different label for classifying the given data.
|