Techniques for implementing a performance-adaptive sampling strategy towards fast and accurate graph neural networks are provided. In one technique, a graph that comprises multiple nodes and edges connecting the nodes is stored. An embedding for each node is initialized, as well as a sampling policy for sampling neighbors of nodes. One or more machine learning techniques are used to train a graph neural network and learn embeddings for the nodes. Using the one or more machine learning techniques comprises, for each node: (1) selecting, based on the sampling policy, a set of neighbors of the node; (2) based on the graph neural network and embeddings for the node and the set of neighbors, computing a performance loss; and (3) based on a gradient of the performance loss, modifying the sampling policy.

In some instances, the disclosure provides a method for identifying access anomalies using network graphs. The method comprises obtaining access data for an entity, generating a network graph baseline profile based on the plurality of data elements, generating a network graph current profile based on the plurality of data elements, generating comparison data based on comparing the plurality of baseline network graphs with the one or more current network graphs and comparing the plurality of baseline nodes and the plurality of baseline edges with the plurality of current nodes and the plurality of current edges, determining, based on the comparison data, anomaly data comprising one or more flagged network accesses to the enterprise system, and providing the anomaly data indicating the flagged network accesses to an authentication system.

Disclosed is a two-stage frequency selection method and device for microwave frequency sweep data. The method includes: acquiring microwave frequency sweep data; performing frequency selection on the microwave frequency sweep data by using a random forest-recursive feature elimination algorithm, taking a preset parameter in the random forest-recursive feature elimination algorithm as a hyper-parameter, changing the value of the hyper-parameter, and generating a series of candidate frequency subsets within different frequencies; building prediction models on the basis of the frequency sweep data corresponding to the candidate frequency subsets of different frequencies; evaluating the performance of each prediction model by means of 10 fold cross validation, and calculating evaluation index values of model performance; and taking the evaluation indexes as a voting basis, and selecting an optimal frequency subset by using a majority voting method.

A computer-implemented method for calculating Scalable Attributed Graph Embedding for Large-Scale Graph Analytics that includes computing a node embedding for a first node-attributed graph in a node embedded space. One or more random attributed graphs is generated in the node embedded space. A graph embedding operation is performed using a dissimilarity measure between one or more raw graphs and the one or more generated random graphs, and an edge-attributed graph into a second node-attributed graph using an adjoint graph.

The present disclosure provides systems and methods for predicting a disease state of a subject using ultrasound imaging and ancillary information to the ultrasound imaging. At least two quantitative measurements of a subject, including at least one measurement taken using ultrasound imaging, as part of quantified information can be identified. One of the quantitative measurements can be compared to a first predetermined standard, included as part of ancillary information to the quantified information, in order to identify a first initial value. Further, another of the quantitative measurements can be compared to a second predetermined standard, included as part of the ancillary information, in order to identify a second initial value. Subsequently, the quantitative information can be correlated with the ancillary information using the first initial value and the second initial value to determine a final value that is predictive of a disease state of the subject.

In exemplary aspects of optimizing data centers, historical data corresponding to a data center is collected. The data center includes a plurality of systems. A data center representation is generated. The data center representation can be one or more of a schematic and a collection of data from among the historical data. The data center representation is encoded into a neural network model. The neural network model is trained using at least a portion of the historical data. The trained model is deployed using a first set of inputs, causing the model to generate one or more output values for managing or optimizing the data center with respect to design and risk aspects.

A method may include generating views materializing tensors generated by a convolutional neural network operating on an image. Determining the outputs of the convolutional neural network operating on the image with a patch occluding various portions of the image. The outputs being determined by generating queries on the views that performs, based at least on the changes associated with occluding different portions of the image, partial re-computations of the views. A heatmap may be generated based on the outputs of the convolutional neural network. The heatmap may indicate the quantities to which the different portions of the image contribute to the output of the convolutional neural network operating on the image. Related systems and articles of manufacture, including computer program products, are also provided.

A method for creating a machine learning system. The method includes: providing a directed graph including an input node and an output node, a probability being in each case assigned to each edge which characterizes the probability with which an edge is drawn. The probabilities are manipulated as a function of a characteristic degree of an exploration of the architectures of the directed graph prior to a random drawing of the architectures.

A computer-implemented method is provided that includes learning causal relationships between two or more application micro-services, and applying the learned causal relationships to dynamically localize an application fault. First micro-service error log data corresponding to selectively injected errors is collected. A learned causal graph is generated based on the collected first micro-service error log data. Second micro-service error log data corresponding to a detected application and an ancestral matrix is built using the learned causal graph and the second micro-service error log data. The ancestral matrix is leveraged to identify the source of the error, and the micro-service associated with the identified error source is also subject to identification. A computer system and a computer program product are also provided.

The present disclosure relates to neural network model processing methods. One example method includes obtaining an operation process of a neural network model, where the operation process is represented by at least one first-type operator and a plurality of second-type operators, and obtaining a first computation graph of the neural network model based on the operation process. In the operation process, the first-type operator includes a boundary identifier, and computational logic of the first-type operator is represented by a group of second-type operators. For any first-type operator, a range of second-type operators included in the any first-type operator is indicated by a boundary identifier in the any first-type operator.