Abstract: A guided sampling tool guides the sampling of datapoints in large datasets. Generally, the guided sampling tool applies a machine learning algorithm to a database of historical issues encountered by an organization to guide the sampling of a large dataset. The guided sampling tool can evaluate and change provided variables and weights for performing a sampling. After the datapoints are sampled, the guided sampling tool compares the historic transactions represented by those datapoints to baseline images to determine if the historic transactions encountered a problem or issue, which would affect the overall quality assessment.

Abstract: A device for removing a noise artifact from a document receives a scan of the document, where the document contains a noise artifact at least partially obstructing a portion of the document. The device generates an image of the document, and extracts a first set of features from the image. The device identifies noise artifact features from the first set of features, and generates a second set of features by removing the noise artifact features. The device generates a test clean image of the document based on the second set of features. The device determines whether a portion of the test clean image that previously displayed the noise artifact corresponds to a counterpart portion of the training clean image. If it is determined that the portion of the test clean image corresponds to the counterpart portion of the training clean image, the device outputs the test clean image.

Abstract: An apparatus includes a memory and processor. The memory stores a graphical representation of a first database that includes source tables and columns. The graphical representation includes nodes associated with source tables and columns. The processor receives an instruction to transfer data from the first database to columns of a second database specified by destination column names. The processor identifies a subset of source columns such that a similarity score for each is greater than a threshold. The similarity score indicates a degree of similarity between a specific destination column name and either the assigned source column name or a related term. The processor uses the subset to generate a pruned graphical representation that includes a subset of nodes. The processor uses this to generate executable code configured to copy data from the first database, determined from the pruned graphical representation, into the second database.

Abstract: Systems, computer program products, and methods are described herein for implementing cognitive self-healing in knowledge based deep learning models. The present invention is configured to receive, via the real-time resource transmission session, one or more query strings from the user; transform the one or more query strings into one or more multi-dimensional query vectors; electronically retrieve one or more multi-dimensional resource vectors from a resource repository; determine a similarity index between the one or more multi-dimensional query vectors and one or more multi-dimensional resource vectors; determine a first multi-dimensional resource vector associated with at least one of the one or more multi-dimensional resource vectors; rasterize the first multi-dimensional resource vector into one or more grid of pixels to generate a first resource; and display, via the real-time resource transmission session, the first resource.

Abstract: Evolutionary learning techniques are used to validate and prioritize open source software libraries for subsequently determining the best open source software library for a specified technical project. Data associated with the open source software candidates is collected into a cluster and, at an eligibility layer, a fitness score is determined for each of the open source software candidate. Candidates that are determined to meet a required fitness score threshold are passed to the crossover layer, at which, software and hardware standards rules are applied to the open source software metadata to validate the open source software. Invalid candidates are held in queue and subjected to rework analysis. A mutation layer executes the crossover layer iteratively until a predetermined volume of open source candidates results. A ranking layer provides a prioritized ranking list, based on the fitness score, of those open source software candidates that have validated.

Abstract: A software code optimizer automatically detects inefficiencies in software code and corrects them. Generally, the software code optimizer converts software code into a graph representing the workflows and relationships in the software code. The graph is then converted into vectors that represent each workflow in the software code. The vectors are assembled into a matrix that represents the software code. The matrix may be stored in a cluster in a database as an example of optimized software code or be compared with other matrices stored as clusters in the database to determine whether the software code is optimized. The software code optimizer can change the software code to be more efficient if a matrix for an optimized version of the software code is found in the database.

Abstract: Systems, computer program products, and methods are described herein for preserving image and acoustic sensitivity using reinforcement learning. The present invention is configured to initiate a file editing engine on the audiovisual file to separate the audiovisual file into a video component and an audio component; initiate a convolutional neural network (CNN) algorithm on the video component to identify one or more sensitive portions in the one or more image frames; initiate an audio word2vec algorithm on the audio component to identify one or more sensitive portions in the audio component; initiate a masking algorithm on the one or more image frames and the audio component; generate a masked video component and a masked audio component based on at least implementing the masking action policy; and bind, using the file editing engine, the masked video component and the masked audio component to generate a masked audiovisual file.

Abstract: Systems, computer program products, and methods are described herein for immersive deep learning in a virtual reality environment.

Abstract: Systems, computer program products, and methods are described herein for enhanced data security in a virtual reality environment. The present invention is configured to initiate an adversarial testing protocol on the virtual reality platform; transmit one or more adversarial inputs to one or more trained machine learning models associated with the virtual reality platform; receive one or more predicted class labels for the one or more adversarial inputs; compare the one or more predicted class labels with one or more observed class labels associated with the one or more adversarial inputs; determine an error rate of the one or more trained machine learning models; determine a first exposure score for the virtual reality platform based on at least the error rate; transmit control signals configured to cause the computing device associated with the user to display the first exposure score for the virtual reality platform.

Abstract: An apparatus includes a memory and processor. The memory stores a graphical representation of a first database that includes source tables and columns. The graphical representation includes nodes associated with source tables and columns. The processor receives an instruction to transfer data from the first database to columns of a second database specified by destination column names. The processor identifies a subset of source columns such that a similarity score for each is greater than a threshold. The similarity score indicates a degree of similarity between a specific destination column name and either the assigned source column name or a related term. The processor uses the subset to generate a pruned graphical representation that includes a subset of nodes. The processor uses this to generate executable code configured to copy data from the first database, determined from the pruned graphical representation, into the second database.

Abstract: A system for administering iterations in a continuous software development lifecycle framework is provided. In particular, the system may be configured to gather one or more requirements from one or more entity systems associated with an entity, gather information associated with the one or more requirements and one or more users from the one or more entity systems, wherein the one or more users are associated with the entity, process the one or more requirements and the information via a natural language processing engine to generate transformed information, identify one or more conflicts associated with the one or more requirements, identify one or more exposures associated with the one or more requirements, and generate one or more decisions based on the transformed information, the one or more conflicts, and the one or more exposures.

Abstract: An analysis tool receives an image as an input matrix. A first convolutional kernel is determined by performing exclusive nor operations between the input matrix and a first weight vector. A first binary kernel is determined based on the first convolutional kernel. A first layer feature map is determined by convoluting the input matrix using the first binary kernel. A second convolutional kernel is determined by performing exclusive nor operations between the first layer feature map and the second weight vector. A pooled kernel is determined based on the second convolutional kernel. A second binary kernel is determined, based on the pooled kernel. A second layer feature map is determined by convoluting the first layer feature map using the second binary kernel. A probability is determined that the input matrix is associated with a predetermined class of images. If the probability is greater than a threshold, classification results are provided.

Abstract: Systems, computer program products, and methods are described herein for immersive deep learning in a virtual reality environment. The present invention is configured to electronically receive, via the extended reality platform, an image of a financial resource; electronically receive, via the extended reality platform, a first user input selecting a machine learning model type; electronically receive, via the extended reality platform, a second user input selecting one or more interaction options; initiate a machine learning model on the image; extract, using the machine learning model, one or more features associated with the image; generate, using the saliency map generator, a saliency map for the image by superimposing the one or more features on the image; and transmit control signals configured to cause the computing device associated with the user to display, via the extended reality platform, the saliency map associated with the image.

Abstract: An adaptive gamified portal builder is provided. The portal builder may include a camera that captures video input and a microphone that captures audio input. The portal builder may identify affective features from the captured audio and video inputs. Using generative adversarial networks (GAN), the portal builder may generate a user interface (UI) output based on the affective features. Using a generator neural network, the portal builder may generate an image. Using a discriminator neural network, the portal builder may access a repository of gamified portal features associated with the cluster of affective vectors and validate the image. The portal builder may iterate through the generator network and the discriminator network to enhance the image. The portal builder may generate a UI image output comprising the enhanced image. Using a UI encoder, the portal builder may generate source code for the UI image output.

Abstract: A tool uses a graph-based approach to analyze scripts to determine whether the scripts pose security threats when executed. The tool breaks down scripts into component steps and generates a graph based on those steps. The tool then converts the graph into a vector and compares that vector with clusters of other vectors. Based on that comparison, the tool determines whether the script will cause a security vulnerability. If the script causes a security threat when executed, the script may be prevented from executing.

Abstract: A software code optimizer automatically detects inefficiencies in software code and corrects them. Generally, the software code optimizer converts software code into a graph representing the workflows and relationships in the software code. The graph is then converted into vectors that represent each workflow in the software code. The vectors are assembled into a matrix that represents the software code. The matrix may be stored in a cluster in a database as an example of optimized software code or be compared with other matrices stored as clusters in the database to determine whether the software code is optimized. The software code optimizer can change the software code to be more efficient if a matrix for an optimized version of the software code is found in the database.

Abstract: A system accesses a disk image of a first software container and collected sensor data for a computer server. The system sequentially analyzes the sequence of layers of the disk image, and generates, based on the sequential analysis of the sequence of layers of the disk image, an auto coding sequence. The auto coding sequence includes a sequence of instructions for creating a new disk image. The system determines, based on the collected sensor data and the sequential analysis of the sequence of layers of the disk image, a sequential list of software needed for the computer server. The system determines, using the sequential list of software needed for the computer server, a plurality of infra requirements for a new software container. The system generates the new software container and the new disk image using the auto coding sequence and the plurality of infra requirements.

Abstract: A system includes a production server, a backup server, a telemetry analyzer, a memory, and a hardware processor. The telemetry analyzer takes snapshots of various performance metrics of the production server. The memory stores a log of previous disasters that occurred on the production server. The log includes a snapshot of the production server performance metrics from the time each disaster occurred. The memory also stores recovery scripts for each logged disaster. Each script provides instructions for resolving the linked disaster. The hardware processor uses a machine learning architecture to train an autoencoder. The trained autoencoder receives new snapshots from the telemetry analyzer and generates a reconstruction of the new snapshots. The hardware processor then determines a threshold for distinguishing between server disasters and minor anomalies. This distinction is made by comparing the difference between the reconstruction of the new snapshots and the new snapshots with the threshold.

Abstract: A language proficiency analyzer automatically evaluates a person's language proficiency by analyzing that person's oral communications with another person. The analyzer first enhances the quality of an audio recording of a conversation between the two people using a neural network that automatically detects loss features in the audio and adds those loss features back into the audio. The analyzer then performs a textual and audio analysis on the improved audio. Through textual analysis, the analyzer uses a multi-attention network to determine how focused one person is on the other and how pleased one person is with the other. Through audio analysis, the analyzer uses a neural network to determine how well one person pronounced words during the conversation.

Abstract: Aspects of the disclosure relate to a machine learning based decentralized business planning system. A computing platform may identify an event likely to impact one or more business operations. Subsequently, the computing platform may receive, for the event, data from one or more sources of data. Then, the computing platform may generate a data structure including a plurality of nodes, where the plurality of nodes corresponds to the received data. Then, the computing platform may authenticate, by utilizing a distributed ledger-based verification system, the plurality of nodes. Subsequently, the computing platform may perform, based on business rules applicable to the one or more business operations, analytics on the plurality of authenticated nodes. Then, the computing platform may generate, based on the analytics, a contingency plan to mitigate the impact to the one or more business operations, and may provide, via an interactive graphical user interface, the contingency plan.