Methods, systems, and computer program products for determining data representative of bias within a model are provided herein. A computer-implemented method includes obtaining a first dataset on which a model was trained, wherein the first dataset contains protected attributes, and a second dataset on which the model was trained, wherein the protected attributes have been removed from the second dataset; identifying, for each of the one or more protected attributes in the first dataset, one or more attributes in the second dataset correlated therewith; determining bias among at least a portion of the identified correlated attributes; and outputting, to at least one user, identifying information pertaining to the one or more instances of bias.

A method and a device for improving performance of a data processing model, a storage medium and an electronic device are provided. A piece of data in a determined test data read currently is determined as target data. Outlier detection parameters in a detection module are acquired. Detection of concept drift is performed on the data processing model based on the target data and the outlier detection parameters. A detection module is triggered to update each of the outlier detection parameters and the data processing model is retrained when concept drift is successfully detected. After the data processing model is already trained, a piece of data to be read next is determined as the target data, the updated outlier detection parameters in the detection module are acquired, and the detection is resumed until all the pieces of data in the test data stream are read.

A method and a device for improving performance of a data processing model, a storage medium and an electronic device are provided. A piece of data in a determined test data read currently is determined as target data. Outlier detection parameters in a detection module are acquired. Detection of concept drift is performed on the data processing model based on the target data and the outlier detection parameters. A detection module is triggered to update each of the outlier detection parameters and the data processing model is retrained when concept drift is successfully detected. After the data processing model is already trained, a piece of data to be read next is determined as the target data, the updated outlier detection parameters in the detection module are acquired, and the detection is resumed until all the pieces of data in the test data stream are read.

A device, method and system for pruning video analytics parameters is provided. A device analyzes images from a camera installed at an installation location using a full initial machine learning model trained to detect objects in the images using video analytics parameters, the full initial machine learning model being independent of the installation location. The device gathers, for a period of time, statistics for: frequency of use of the video analytics parameters and/or error generation frequency in detecting the objects in the images. After the period of time, based on the statistics, the device determines that given video analytics parameters meets one or more pruning conditions; and, in response, prunes the given video analytics parameters such that the full initial machine learning model, initially provided for the camera, changes to an updated machine learning model for the camera.

A computer-implemented method for fusing an end-to-end speech recognition model with an external language model (ExternalLM) is provided. The method includes obtaining an output of the end-to-end speech recognition model. The output is a probability distribution. The method further includes transforming, by a hardware processor, the probability distribution into a transformed probability distribution to relax a sharpness of the probability distribution. The method also includes fusing the transformed probability distribution and a probability distribution of the ExternalLM for decoding speech.

Systems, apparatuses, and methods are provided herein for providing natural language explanation to black-box algorithm generated outcome. The system is configured to determine a regression coefficient for each of the plurality of attributes based regression analysis, determine a decision tree based on the input data and the output data and a decision path of a select data item in the decision tree, generate natural language explanation of a categorization of the select data item based on the relevant attributes and regression coefficients associated with each of the relevant attributes, wherein the natural language explanation identifies at least one relevant attribute and an effect of the at least one relevant attribute of the data item on the categorization, and transmit to a user interface device for display, the categorization of the select data item along with the natural language explanation of the categorization of the select data.

Embodiments of the present invention provide a system for an intelligent and adaptive real time valuation engine. The described system receives a plurality of input data from data pipelines, the input data including a request for an available transaction value given certain parameters. The system provides the plurality of input data into a long short term memory neural network engine for the identification of dependencies between data characteristics and transaction values, providing a continuous stream of predicted transaction for multiple nodes of the neural network engine. The system then assigns weighting to the outputs of the neural network engine, and determines the available transaction values based on these weighted outputs.

In one aspect, the present disclosure relates to a method for or performing single-pass object detection and image classification. The method comprises receiving image data for an image in a system comprising a convolutional neural network (CNN), the CNN comprising a first convolutional layer, a last convolutional layer, and a fully connected layer; providing the image data to an input of the first convolutional layer; extracting multi-channel data from the output of the last convolutional layer; and summing the extracted data to generate a general activation map; and detecting a location of an object within the image by applying the general activation map to the image data.

In one aspect, the present disclosure relates to a method for or performing single-pass object detection and image classification. The method comprises receiving image data for an image in a system comprising a convolutional neural network (CNN), the CNN comprising a first convolutional layer, a last convolutional layer, and a fully connected layer; providing the image data to an input of the first convolutional layer; extracting multi-channel data from the output of the last convolutional layer; and summing the extracted data to generate a general activation map; and detecting a location of an object within the image by applying the general activation map to the image data.

Systems and methods for identifying writing systems utilized in documents. An example method comprises: receiving a document image; splitting the document image into a plurality of image fragments; generating, by a neural network processing the plurality of image fragments, a plurality of probability vectors, wherein each probability vector of the plurality of probability vectors is produced by processing a corresponding image fragments and contains a plurality of numeric elements, and wherein each numeric element of the plurality of numeric elements reflects a probability of the image fragment containing a text associated with a respective writing system; computing an aggregated probability vector by aggregating the plurality of probability vectors, wherein each numeric element of the aggregated probability vector reflects a probability of the image containing a text associated with a writing system that is identified by an index of the numeric element within the aggregated probability vector; and responsive to determining that a maximum numeric element of the aggregated probability vector exceeds a predefined threshold value, concluding that the document image contains one or more symbols associated with a respective writing system.