Back to results
US 12,572,626 B2Granted
Method and system for optimizing a pair of affine classifiers based on a diversity metric
AssigneeGenesee Valley Innovations, LLC
InventorsShantanu Rane, Bashir Sadeghi, Alejandro E. Brito
Life of this patent
Estimating the term…
Abstract
One embodiment provides a method and system which facilitates optimizing a pair of affine classifiers based on a diversity metric. During operation, the system defines a diversity metric based on an angle between decision boundaries of a pair of affine classifiers. The system includes the diversity metric as a regularization term in a loss function optimization for designing the pair of affine classifiers, wherein the designed pair of affine classifiers are mutually orthogonal. The system predicts an outcome for a testing data object based on the designed pair of mutually orthogonal affine classifiers.
Claims20
A computer-executable method, comprising:
- determining data objects, wherein a data object comprises a training data object or a testing data object;
- defining a diversity metric based on an angle between decision boundaries of a pair of affine classifiers;
- including the diversity metric as a regularization term in a loss function optimization for designing the pair of affine classifiers, wherein the designed pair of affine classifiers are mutually orthogonal;
- predicting an outcome or class associated with a respective data object based on the designed pair of affine classifiers;
- displaying, on a device associated with a user, information related to the predicted outcome or class associated with the respective data object, wherein the information comprises one or more of:
- a type of the respective data object, including whether the respective data object is clean or modified based on an adversarial attack;
- a type associated with the adversarial attack, in response to the respective data object being modified based on an adversarial attack;
- an approach associated with the respective data object, including whether the respective data object is associated with diversity;
- and
- a result based on a respective classifier of the pair of affine classifiers;
- receiving, based on an interaction by the user with the displayed information, a request to change a configuration or setting related to one or more of the type of the respective data object, the type associated with an adversarial attack, the approach associated with the respective data object, the regularization term, and a selected classifier of the pair of affine classifiers;
- predicting an updated outcome or class associated with the respective data object based on the changed configuration or setting, wherein the respective data object comprises attacked data modified based on an adversarial technique;
- and
- displaying, on the device associated with the user, the information related to the predicted updated outcome or class associated with the respective data object.
11 more elements
A computer system for facilitating data classification, the computer system comprising:
- a processor;
- and
- a storage device storing instructions that when executed by the processor cause the processor to perform a method, the method comprising:
- determining data objects, wherein a data object comprises a training data object or a testing data object;
- defining a diversity metric based on an angle between decision boundaries of a pair of affine classifiers;
- including the diversity metric as a regularization term in a loss function optimization for designing the pair of affine classifiers, wherein the designed pair of affine classifiers are mutually orthogonal;
- predicting an outcome or class associated with a respective data object based on the designed pair of mutually orthogonal affine classifiers;
- displaying, on a device associated with a user, information related to the predicted outcome or class associated with the respective data object, wherein the information comprises one or more of:
- a type of the respective data object, including whether the respective data object is clean or modified based on an adversarial attack;
- a type associated with the adversarial attack, in response to the respective data object being modified based on an adversarial attack;
- an approach associated with the respective data object, including whether the respective data object is associated with diversity;
- and
- a result based on a respective classifier of the pair of affine classifiers;
- receiving, based on an interaction by the user with the displayed information, a request to change a configuration or setting related to one or more of the type of the respective data object, the type associated with an adversarial attack, the approach associated with the respective data object, the regularization term, and a selected classifier of the pair of affine classifiers;
- predicting an updated outcome or class associated with the respective data object based on the changed configuration or setting, wherein the respective data object comprises attacked data modified based on an adversarial technique;
- and
- displaying, on the device associated with the user, the information related to the predicted updated outcome or class associated with the respective data object.
14 more elements
A non-transitory computer-readable storage medium storing instructions that when executed by a computer cause the computer to perform a method, the method comprising:
- determining data objects, wherein a data object comprises a training data object or a testing data object;
- defining a diversity metric based on an angle between decision boundaries of a pair of affine classifiers;
- including the diversity metric as a regularization term in a loss function optimization for designing the pair of affine classifiers, wherein the designed pair of affine classifiers are mutually orthogonal;
- predicting an outcome or class associated with a respective data object based on the designed pair of mutually orthogonal affine classifiers;
- displaying, on a device associated with a user, information related to the predicted outcome or class associated with the respective data object, wherein the information comprises one or more of:
- a type of the respective data object, including whether the respective data object is clean or modified based on an adversarial attack;
- a type associated with the adversarial attack, in response to the respective data object being modified based on an adversarial attack;
- an approach associated with the respective data object, including whether the respective data object is associated with diversity;
- and
- a result based on a respective classifier of the pair of affine classifiers;
- receiving, based on an interaction by the user with the displayed information, a request to change a configuration or setting related to one or more of the type of the respective data object, the type associated with an adversarial attack, the approach associated with the respective data object, the regularization term, and a selected classifier of the pair of affine classifiers;
- predicting an updated outcome or class associated with the respective data object based on the changed configuration or setting, wherein the respective data object comprises attacked data modified based on an adversarial technique;
- and
- displaying, on the device associated with the user, the information related to the predicted updated outcome or class associated with the respective data object.
11 more elements
Dependencies are inferred from the claim text.