Abstract: A system and method includes generating approximate distributions for distinct classes of data samples; computing a first partial Jensen-Shannon (JS) divergence and a second partial JS divergence based on the approximate distribution of the disparity affected class of data samples with reference to the approximate distribution of the control class of data samples; computing a disparity divergence based on the first partial JS divergence and the second partial JS divergence; generating a distribution-matching term based on the disparity divergence, wherein the distribution-matching term mitigates an inferential disparity between the control class of data samples and the disparity affected class of data samples during a training of an unconstrained artificial neural network; constructing a disparity-constrained loss function based on augmenting a target loss function with the distribution-matching term; and transforming the unconstrained ANN to a disparity-constrained ANN based on a training of the unconstraine

Abstract: Systems and methods for training models to improve fairness.

Abstract: A system and method includes generating approximate distributions for distinct classes of data samples; computing a first partial Jensen-Shannon (JS) divergence and a second partial JS divergence based on the approximate distribution of the disparity affected class of data samples with reference to the approximate distribution of the control class of data samples; computing a disparity divergence based on the first partial JS divergence and the second partial JS divergence; generating a distribution-matching term based on the disparity divergence, wherein the distribution-matching term mitigates an inferential disparity between the control class of data samples and the disparity affected class of data samples during a training of an unconstrained artificial neural network; constructing a disparity-constrained loss function based on augmenting a target loss function with the distribution-matching term; and transforming the unconstrained ANN to a disparity-constrained ANN based on a training of the unconstraine

Abstract: Methods, non-transitory computer readable media, and model evaluations systems for understanding diverse machine learning models (MLMs) are disclosed. In some examples, a feature contribution value is determined for features included in a reference or evaluation input data set. The evaluation input data set represents a protected class population and each feature contribution value identifies a contribution by a feature to a difference in output generated by an MLM for the evaluation input data set. Model explanation information is generated using the feature contribution values and execution of the MLM is monitored. The model explanation information explains the difference in output generated by the MLM for the evaluation input data set and includes information relating to a model-based decision. A report is generated from a knowledge graph for the MLM and output via a GUI to an operator device that includes the model explanation information.

Abstract: Systems and methods for training models to improve fairness.

Abstract: Systems and methods for generating and processing modeling workflows.

Abstract: Systems and methods for understanding diverse machine learning models.

Abstract: A system and method includes generating approximate distributions for distinct classes of data samples; computing a first partial Jensen-Shannon (JS) divergence and a second partial JS divergence based on the approximate distribution of the disparity affected class of data samples with reference to the approximate distribution of the control class of data samples; computing a disparity divergence based on the first partial JS divergence and the second partial JS divergence; generating a distribution-matching term based on the disparity divergence, wherein the distribution-matching term mitigates an inferential disparity between the control class of data samples and the disparity affected class of data samples during a training of an unconstrained artificial neural network; constructing a disparity-constrained loss function based on augmenting a target loss function with the distribution-matching term; and transforming the unconstrained ANN to a disparity-constrained ANN based on a training of the unconstraine

Abstract: A system and method includes generating approximate distributions for distinct classes of data samples; computing a first partial Jensen-Shannon (JS) divergence and a second partial JS divergence based on the approximate distribution of the disparity affected class of data samples with reference to the approximate distribution of the control class of data samples; computing a disparity divergence based on the first partial JS divergence and the second partial JS divergence; generating a distribution-matching term based on the disparity divergence, wherein the distribution-matching term mitigates an inferential disparity between the control class of data samples and the disparity affected class of data samples during a training of an unconstrained artificial neural network; constructing a disparity-constrained loss function based on augmenting a target loss function with the distribution-matching term; and transforming the unconstrained ANN to a disparity-constrained ANN based on a training of the unconstraine

Abstract: A system and method includes generating approximate distributions for distinct classes of data samples; computing a first partial Jensen-Shannon (JS) divergence and a second partial JS divergence based on the approximate distribution of the disparity affected class of data samples with reference to the approximate distribution of the control class of data samples; computing a disparity divergence based on the first partial JS divergence and the second partial JS divergence; generating a distribution-matching term based on the disparity divergence, wherein the distribution-matching term mitigates an inferential disparity between the control class of data samples and the disparity affected class of data samples during a training of an unconstrained artificial neural network; constructing a disparity-constrained loss function based on augmenting a target loss function with the distribution-matching term; and transforming the unconstrained ANN to a disparity-constrained ANN based on a training of the unconstraine

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for machine learning model fairness and explainability. In some implementations, a method includes obtaining data relating to a plurality of potential borrowers; providing the data to the trained machine learning model; obtaining, by the trained machine learning model’s processing of the provided data, the one or more outputs of the trained machine learning model; and automatically generating a report that explains the one or more outputs of the trained machine learning model with respect to one or more fairness metrics and one or more accuracy metrics; and providing the automatically generated report for display on a user device.

Abstract: Systems and methods for training models to improve fairness.

Abstract: Systems and methods for training models to improve fairness.

Abstract: Systems and methods for training models to improve fairness.

Abstract: Systems and methods for understanding diverse machine learning models.

Abstract: Systems and methods for explaining non-differentiable models and differentiable models.

Abstract: Systems and methods for generating and processing modeling workflows.

Abstract: The subject invention can track and apply user edits to a source document as a sequence of changes. The changes can be applied in a document or spatial order irrespective of temporal factors. The invention can maintain intervals that represent user operations (e.g., insertions, deletions, zero-net-length changes). As well, the invention can infer a location in the original document that corresponds to a particular operation. In accordance therewith, the invention can arrange temporally sequenced user document modifications into an order consistent with the layout of the document file encoding. This functionality of mapping re-sequenced changes into the original document data representation is one novel feature of the invention. The invention can enable portions of the source document loaded into memory on an as-needed basis whereby changes relevant to the instant portion can be made.

Abstract: A system management framework is described for application delivery and configuration management of mobile devices. The framework includes a management server and a mobile computing device. The management server is configured to communicate download instructions for purposes of configuration management of mobile computing devices. The mobile computing device is configured to connect to the management server over a non-persistent connection. The mobile computing device requests download instructions from the management server to determine any offerings that may be available for download and installation by the mobile computing device. Any offerings presented by the management server represent one or more files that have been made available since a last successful download operation conducted by the mobile computing device. The mobile computing device allows a user to accept or reject download and installation of any one or more of the offerings.

Abstract: An animation system provides synchronization services to synchronize actions of two more interactive user interface characters that are displayed simultaneously. The animation services allow applications to make animation requests to control the actions of characters on the display. These actions include playing one of the character's animation sequences and generating speech output with lip-synched animation of the character's mouth. Accessible via script commands or an Application Programming Interface, the synchronization services allow an application to control interaction between two or more characters on the display. Applications can synchronize actions by invoking straightforward commands such as Wait, Interrupt, or Stop.