Abstract: An exemplary device for classifying an image includes a receiving unit that receives image data. The device also includes a hardware processor including a neural network architecture to extract a plurality of features from the image data, filter each feature extracted from the image data, concatenate the plurality of filtered features to form an image vector, evaluate the plurality of concatenated features in first and second layers of a plurality of fully connected layers of the neural network architecture based on an amount of deviation in the features determined at each fully connected layer, and generate a data signal based on an output of the plurality of fully connected layers. A transmitting unit sends the data signal to a peripheral or remote device.

Abstract: A method and system for explaining a decision process of a machine learning model that includes inputting into a machine learning model a first input data file; receiving a first output data file from the machine learning model based on the first input data file; executing an adversarial attack on the machine learning model, creating a mapping of the one or more units of data of the first input data file with changes by the adversarial attack exceeding a first threshold to one or more segments of the first input data file; determining a density of the changes to the one or more units of data in each of the one or more segments; and displaying the one or more segments of the first input data file having a density of changes to the one or more units of data exceeding a second threshold via a graphical user interface.

Abstract: Exemplary systems and methods to extract, transform, and save to memory features from a training and a test dataset at extraction layers in a machine-learning model. For each data element in the training dataset, at each extraction layer: feature maps are created and grouped by k unique data labels to construct a set of k class-conditional distributions. For each data element in the datasets: distance sets between each feature map of each extraction layer and the extraction layer's class-conditional distributions are calculated and reduced to distance summary metrics. A drift test statistic for each extraction layer is computed between the datasets by comparing the extraction layer's distance summary metric distributions of the test dataset to distance summary metric distributions of the training dataset. The measure of drift between the datasets is computed by combining the test statistics of the extraction layers through a mathematical transform.

Abstract: A method and system for explaining a decision process of a machine learning model that includes inputting into a machine learning model a first input data file; receiving a first output data file from the machine learning model based on the first input data file; executing an adversarial attack on the machine learning model, creating a mapping of the one or more units of data of the first input data file with changes by the adversarial attack exceeding a first threshold to one or more segments of the first input data file; determining a density of the changes to the one or more units of data in each of the one or more segments; and displaying the one or more segments of the first input data file having a density of changes to the one or more units of data exceeding a second threshold via a graphical user interface.

Abstract: A computer-implemented method for training a neural network system. The method includes receiving at least a first data vector at a first layer of the neural network system; applying a function to the first data vector to generate at least a second data vector, wherein the function is based on a layer parameter of the first layer that includes at least a weight matrix of the first layer; comparing at least the first data vector and the second data vector to obtain a loss value that represents a difference between the first data vector and the second data vector; updating the layer parameter based on the loss value; and adjusting the layer parameter based on a comparison of the updated layer parameter with a threshold value of the first layer.

Abstract: An exemplary device for classifying an image includes a receiving unit that receives image data. The device also includes a hardware processor including a neural network architecture to extract a plurality of features from the image data, filter each feature extracted from the image data, concatenate the plurality of filtered features to form an image vector, evaluate the plurality of concatenated features in first and second layers of a plurality of fully connected layers of the neural network architecture based on an amount of deviation in the features determined at each fully connected layer, and generate a data signal based on an output of the plurality of fully connected layers. A transmitting unit sends the data signal to a peripheral or remote device.

Abstract: An exemplary device for classifying an image includes a receiving unit that receives image data. The device also includes a hardware processor including a neural network architecture to extract a plurality of features from the image data, filter each feature extracted from the image data, concatenate the plurality of filtered features to form an image vector, evaluate the plurality of concatenated features in first and second layers of a plurality of fully connected layers of the neural network architecture based on an amount of deviation in the features determined at each fully connected layer, and generate a data signal based on an output of the plurality of fully connected layers. A transmitting unit sends the data signal to a peripheral or remote device.

Abstract: A computer-implemented method for training a neural network system. The method includes receiving at least a first data vector at a first layer of the neural network system; applying a function to the first data vector to generate at least a second data vector, wherein the function is based on a layer parameter of the first layer that includes at least a weight matrix of the first layer; comparing at least the first data vector and the second data vector to obtain a loss value that represents a difference between the first data vector and the second data vector; updating the layer parameter based on the loss value; and adjusting the layer parameter based on a comparison of the updated layer parameter with a threshold value of the first layer.