Abstract: Methods and systems are described herein for manipulating a communication acceptance screen, manipulating an interactive communication acceptance icon, and restricting access to accounts based on voice communication parameters. In particular, when a communication is detected from one device to another device, that communication may be risky. Thus, the risk is mitigated by giving a user information about the source of the communication to give a user a chance to reject the communication. In addition, in instances where the user accepts the communication, the system enables monitoring the communication and restricting any accounts that are disclosed within the communication.

Abstract: Various systems and methods are presented regarding applying a non-Clifford logical CCZ gate in a 2D topological code, e.g., a 2D Abelian code. A Kirigami cut and fold process is applied to an initial 3D quantum error-correcting code composed of three copies of 3D surface codes to reduce (cut) a qubit lattice to a hollow cube such that the cube is formed of surfaces comprising qubits. The CCZ is applied to the hollow cube, with the cube subsequently flattened (folded) from the 3D arrangement to a 2D code arrangement. A lattice surgery process can be performed to map/transfer the 2D code arrangement to an ancilla Kirigami code(s), whereupon the 2D code arrangement can be returned to a 2D surface code for further application of a logical Clifford gate, if so desired.

Abstract: As described herein, a system, method, and computer program are provided for creating a new class of instances. A first dataset comprising a plurality of instances is obtained. A first AI engine is used to classify the plurality of instances into a plurality of classes. Each class of instances of the first dataset is associated with a respective action having a positive result or a lack of the positive result. A class is selected and the respective action is executed on instances of the class. Instances of the selected class for which the respective action resulted in a lack of the positive result are collected. The collection of instances are used to train an AI-model, and/or a second AI engine is used to classify the collection of instances, and/or a plurality of artificial instances are synthesized based on the collection of instances.

Abstract: Techniques are disclosed that enable training a global model using gradients provided to a remote system by a set of client devices during a reporting window, where each client device randomly determines a reporting time in the reporting window to provide the gradient to the remote system. Various implementations include each client device determining a corresponding gradient by processing data using a local model stored locally at the client device, where the local model corresponds to the global model.

Abstract: Systems for self-organizing data collection in an industrial environment are disclosed. An example system may include a self-propelled mobile data collector for handling a plurality of sensor inputs from sensors in the industrial environment, wherein the plurality of sensor inputs is configured to sense at least one of: an operational mode, a fault mode, a maintenance mode, or a health status of at least one target system. The system may include a self-organizing system for self-organizing at least one of a storage operation of the data, a data collection operation of the sensors, or a selection operation of the plurality of sensor inputs. The self-organizing system organizes a swarm of self-propelled mobile data collectors to collect data from a plurality of target systems in the industrial environment.

Abstract: Systems and methods for detecting objects in a video are provided. A method can include inputting a video comprising a plurality of frames into an interleaved object detection model comprising a plurality of feature extractor networks and a shared memory layer. For each of one or more frames, the operations can include selecting one of the plurality of feature extractor networks to analyze the one or more frames, analyzing the one or more frames by the selected feature extractor network to determine one or more features of the one or more frames, determining an updated set of features based at least in part on the one or more features and one or more previously extracted features extracted from a previous frame stored in the shared memory layer, and detecting an object in the one or more frames based at least in part on the updated set of features.

Abstract: An edge site maintains a plurality of rendezvous nodes to which a plurality of Internet of Things (IoT) devices are communicatively coupled, where the plurality of rendezvous nodes are also communicatively coupled to an edge hub in the edge site. A cached model is maintained at the edge hub, where the cached model enables the edge site to stay operational in an event of a failure of a cloud hub to which the edge hub is communicatively coupled.

Abstract: A method is provided that includes accessing a multivariate time series of flight data for an aircraft, and iteratively performing runs of genetic programming on groups of the sensors. A population of computer programs is randomly generated from a selected group of the plurality of sensors, and primitive functions selected from a library of primitive functions. The population is iteratively transformed into new generations of the population, and includes sub-rankings of the group of sensors based on a quantitative fitness determined according to selected fitness criterion. A ranking of the group of sensors from the sub-rankings of the group of sensors is produced. An aggregate ranking of the plurality of sensors is produced from the ranking of the group of sensors over a plurality of iterations. And the subset of sensors is selected from the aggregate ranking of the plurality of sensors, and according to selected optimization criterion.

Abstract: A computer-implemented system for processing medical claims is disclosed. The system includes a medical device configured to be manipulated by a user while the user performs a treatment plan; a patient interface associated with the medical device, the patient interface comprising an output configured to present telemedicine information associated with a telemedicine session; and a processor. During the telemedicine session, the processor is configured to receive information from a medical device. Using the device-generated information, the processor is further configured to determine device-based medical coding information. The processor is further configured to transmit the device-based medical coding information to a claim adjudication server.

Abstract: Techniques disclosed herein involve automatically adjusting a control parameter for an operating mode of a medical device. In some embodiments, the techniques involve obtaining data pertaining to a physiological condition of a patient during operation of the medical device. In some embodiments, the techniques further involve determining an optimized value for the control parameter using the data pertaining to the physiological condition of the patient and a cost function, wherein the cost function disproportionately penalizes for an amount of time that a physiological parameter is below a lower bound of a target range and includes a first weighting factor that is dependent on an amount of time that the physiological parameter is outside of the target range, and a second weighting factor, different from the first weighting factor, dependent on the amount of time that the physiological parameter is below the lower bound of the target range.

Abstract: A method for robust identification of a vehicle occupant. Simultaneously acquired data pairs from interior sensors and sensing elements are used as a training data set for global learning of a Deep Canonical Correlation Analysis architecture with two neural networks, wherein an authentication of the respective vehicle occupants from the group of vehicle users is carried out in a first step as an initialization. In a second step or a subsequent use, only one of the two sensing elements is then needed to identify the respective vehicle occupant. Also described is a system, via which the robust identification is provided by way of the method.

Abstract: A network connection control system and method is provided. The network connection control system includes user equipments, base stations, a server and a processing unit. Each base station has a connection range and is configured to connect the user equipment located within the connection range. Each user equipment transmits a network parameter between it and every base station through the base station connected therewith to the server. The server generates a QUBO matrix according to all the network parameters and outputs the QUBO matrix to the processing unit. The processing unit processes the QUBO matrix based on a quantum annealing algorithm to obtain an optimized connection configuration of the user equipments and base stations. The server receives the optimized connection configuration and controls the connection range of each base station accordingly for making an actual connection configuration of the user equipments and base stations identical with the optimized connection configuration.

Abstract: Methods and systems for performing transfer learning with basis scaling and pruning. One method includes obtaining a pre-trained deep convolutional neural network (DCNN), decomposing each weight matrix of the DCNN, and decomposing each convolutional layer by applying the respective decomposed weight matrix to the convolution layer to form a first layer which comprises the left matrix for convolution, and a second layer which comprises the right matrix for convolution. The method also includes providing a basis-scaling convolutional layer having a weight matrix that is derived by a function of singular values and the right singular vectors and training the basis scaling factors of the basis-scaling convolutional layers.

Abstract: In accordance with one or more aspects of the disclosed embodiment, a lift includes at least one load handling device configured so as to reciprocate along a lift axis, the load handling device including a frame forming a payload section with a payload support surface having a common elevation configured to hold one or more pickfaces at the common elevation of the payload support surface, at least one transfer arm movably mounted to the frame, and a drive section connected to the load handling device and being configured to move the load handling device along the lift axis, wherein the one or more pickfaces carried in unison by the payload support surface of the at least one load handling device, define an order sequence of pickfaces on the at least one load handling device according to a predetermined case out order sequence of mixed cases.

Abstract: Apparatus, systems, and methods for analyzing data are described. The data can be analyzed using a hierarchical structure. One such hierarchical structure can comprise a plurality of layers, where each layer performs an analysis on input data and provides an output based on the analysis. The output from lower layers in the hierarchical structure can be provided as inputs to higher layers. In this manner, lower layers can perform a lower level of analysis (e.g., more basic/fundamental analysis), while a higher layer can perform a higher level of analysis (e.g., more complex analysis) using the outputs from one or more lower layers. In an example, the hierarchical structure performs pattern recognition.

Abstract: The systems and methods provide a model deployment criterion. The model deployment criterion indicates a difference in a value against which the proxy model may be measured to determine when, if ever, the proxy model should be deployed to replace the existing rule engine. The model deployment criterion may be keyed to the proxy model (e.g., based on a difference in its size, throughput speed, number of changes, etc.), the existing rule engine (e.g., based on a difference in its age, update occurrences to its rule base, etc.), and/or comparisons between models (e.g., based on differences in results, throughput speed, efficiency, etc.).

Abstract: Obtain, as input, in electronic form, structured information including tasks for a plurality of occupations in a plurality of industries over a length of time; compute, from the structured information, a time series of normalized occupation task shares over the length of time; train a computerized machine learning model, on the time series, to predict future task shares for the plurality of occupations in the plurality of industries; and, with the trained computerized machine learning model, predict the future task shares.

Abstract: According to some embodiments, an available algorithm data store may contain information about a pool of available algorithms. An algorithm selection platform coupled to the available algorithm data store may access the information about the pool of available algorithms and compare the information about each of the pool of available algorithms with at least one requirement associated with the current algorithm executing in the real environment. The algorithm selection platform may then automatically determine algorithm execution context information and, based on said comparison and the algorithm execution context information, select at least one of the pool of available algorithms as a potential replacement algorithm. An indication of the selected at least one potential replacement algorithm may then be transmitted (e.g., to be evaluated in a shadow environment by an algorithm evaluation platform).

Abstract: In one aspect, an example methodology implementing the disclosed techniques includes, by a disk controller of a plurality of disks of a data storage system, collecting usage telemetry from the plurality of disks and initiating determination of one or more candidate time slots for performing a battery learn cycle using a machine learning model. The machine learning model is trained using the usage telemetry from the plurality of disks. The method also includes, by the disk controller, presenting one or more of the one or more candidate time slots for performing a battery learn cycle.

Abstract: Systems and methods are provided to improve traditional chip processing. Using crossbar computations, the convolution layer can be flattened into vectors, and the vectors can be grouped into a matrix where each row or column is a flattened filter. Each submatrix of the input corresponding to a position of a convolution window is also flattened into a vector. The convolution is computed as the dot product of each input vector and the filter matrix. Using intra-crossbar computations, the unused space of the crossbars is used to store replicas of the filters matrices and the unused space in XIN is used to store more elements of the input. In inter-crossbar computations, the unused crossbars are used to store replicas of the filters matrices and the unused XINs are used to store more elements of the input. Then, the method performs multiple convolution iterations in a single step.

Abstract: A system for retrieving patient information using large language models including a computing device configured to receive a natural language query as a function of a user input, input the natural language query into a large language model communicatively connected to the least a processor, receive a computer language query comprising a plurality of nodes from the large language model, map the plurality of nodes to one or more entries in a patient database, receive a database response from the patient database as a function of the mapping, generate a final database query as a function of the database response. query the patient database using the final database query, receive a user response as a function of the final database query, and transmit the user response to a graphical user interface as a function of the final database query.

Abstract: According to an aspect of an embodiment, a method may include obtaining multiple sets of camera images and light detection and ranging (LIDAR) point clouds along a track within a geographic sector of a map. The method may include applying a learning model to the camera images to characterize objects within the camera images within classes of objects to generate segmented images. The method may additionally include mapping the sets of camera images and the LIDAR point clouds to three dimensional points of the geographic sector of the map. The method may also include projecting the three dimensional points onto the segmented images to obtain corresponding classes for the three dimensional points of the geographic sector of the map.

Abstract: Embodiments of the present invention analyze multiple factors—such as user input events, device motion data, other data from the endpoint, or data from an external system (such as a real-time location system)—to make a probabilistic determination whether a walkaway event has occurred.

Abstract: Systems and methods for exchanging synchronization information between processing units using a synchronization network are disclosed. The disclosed systems and methods include a device including a host and associated neural processing units. Each of the neural processing units can include a command communication module and a synchronization communication module. The command communication module can include circuitry for communicating with the host device over a host network. The synchronization communication module can include circuitry enabling communication between neural processing units over a synchronization network. The neural processing units can be configured to each obtain a synchronized update for a machine learning model. This synchronized update can be obtained at least in part by exchanging synchronization information using the synchronization network. The neural processing units can each maintain a version of the machine learning model and can synchronize it using the synchronized update.

Abstract: A determination is made that a mobile device is associated with a reference activity of a user based on motion, orientation, rotational, magnetic field, and/or location data provided by sensors of the device. Activity data associated with the reference activity is obtained from the sensor-provided data. The activity data is recorded on the device for a configured period of time after which it is determined that the device is no longer performing the reference activity. The retained activity data for the reference activity is sent from the device to a network-based behavior analyzer when a network connection is available from the device. The network-based behavior analyzer derives user behaviors for the reference activity based on the activity data.

Abstract: System and method for facilitating engineering of an industrial system (250) are disclosed.

Abstract: A medical image processing apparatus comprising processing circuitry configured to: receive medical image data including a region of interest; generate first rendering data by rendering the medical image data based on a first rendering condition; receive a characteristic value relating to detection certainty with respect to the region of interest by inputting the first rendering data to a machine learning model; determine a property of the first rendering condition based on the characteristic value.

Abstract: Rather than randomly selecting neural networks to classify a media file, the conductor can determine which neural network engines (from the conductor ecosystem of neural networks) are the best candidates to classify a particular portion/segment of the media file (e.g., audio file, image file, video files). The best candidate neural network engine(s) can depend on the nature of the input media and the characteristics of the neural network engines. In object recognition and identification, certain neural networks can classify vehicles better than others, while another group of neural networks can classify structures better. The conductor can take out the guess work and construct in real-time an inter-classifier neural network using one or more layers selected from one or more pre-trained neural network, based on attribute(s) of the media file, to classify the media file.

Abstract: Computer systems and computer-implemented methods train a neural network iteratively training, through machine learning. The iterative training comprises imposing a first is-not-equal-to regularization link between first and second nodes, where imposing the first is-not-equal-to regularization link between the two nodes comprises adding, during back-propagation of partial derivatives through the neural network for a datum in a training data set, a first regularization cost to a network error loss function for the first node that is inversely proportional to a difference between an activation value for the first node for the datum and an activation value for the second node for the datum.

Abstract: Provided herein are systems, methods and computer readable media for managing a sales pipeline, and in some embodiments, calculating supply based on travel propensity. An example method comprises identifying a total demand for a promotion tuple at a geographic location, determining, using a processor, a promotion area for the promotion tuple, the promotion tuple comprising at least a category, price information and a geographic area, identifying one or more promotions offered by a promotion and marketing service that comprise at least the category and the price information of the promotion tuple, determining whether the total demand for the promotion tuple at the geographic location is satisfied, wherein the total demand is satisfied in an instance in which the geographic location is within the promotion area for the one or more promotions, and identifying a demand gap in an instance in which the total demand for the promotion tuple at the geographic location is not satisfied.

Abstract: Systems and methods for cooling a computer environment are disclosed. In at least one embodiment, one or more neural networks can be used to determine one or more temperature control settings associated with one or more servers.

Abstract: A system and method include receiving a first set of variables associated with a real-time request, extracting a predetermined subset of the first set of variables for generating a second set of variables, identifying historical request data, computing a set of parameters based on the first set of variables and the historical request data, generating a plurality of numeric sequences and a plurality of string sequences for the real-time request, converting each of the plurality of string sequences into an encoded string sequence to obtain a plurality of encoded string sequences, inputting the plurality of numeric sequences and the plurality of encoded string sequences into a trained deep machine learning model, and computing a score from the trained deep machine learning model, the score indicative of a likelihood that the real-time request belongs to an unauthorized classification.

Abstract: Techniques for improving system performance based on data characteristics are disclosed. A system may receive updates to a first data set at a first frequency. The system selects a first storage configuration, from a plurality of storage configurations, for storing the first data set based on the first frequency, and stores the first data set in accordance with the first storage configuration. The system may further receive updates to a second data set at a second frequency. The system selects a second storage configuration, from the plurality of storage configurations, for storing the second data set based on the second frequency, and stores the second data set in accordance with the second storage configuration. The second storage configuration is different than the first storage configuration.

Abstract: An approach is presented herein to use an in-situ algorithmic decision methodology during each stage of testing before 2C/4C to decide how long to test, how much margin should be used for each device under the test (DUT) to shorten or eliminate 2C/4C testing. Each DUT will be tested differently based on the risk level or the likelihood of failure at 2C/4C. To be able to achieve this, low-level hardware (HW) based sensors (on the printed circuit board assembly (PCBA), in power module, in silicon components, in silicon component complex, etc.) are used to collect telemetry data with a high frequency data acquisition rate during the testing. As testing is ongoing for each DUT, a margin distribution and algorithm modeling is performed in-situ.

Abstract: A need to retrain a machine learning model is determined based on identification of either a new scenario (i.e., new grouping/population of data variables) or a modification to an existing/trained scenario (i.e., addition and/or deletion of data variables in an already trained scenario). Change occurring across all of the data variables used in training the machine learning model are monitored as opposed to monitoring only the data variables on an individual basis. When concurrent change occurs across multiple inter-related data variables, the need to add or modify a scenario (i.e., retrain a machine learning model) is identified.

Abstract: A machine learning model is sequentially fine-tuned with new training data as the training data becomes available. By using a suitable mix of old and new data, and weighting samples in the training data by age, the model can be efficiently updated to maintain accuracy against a changing malware landscape without manual modifications of the network layers or the computational expense of full retraining.

Abstract: A method for providing an optimal path through a computer network includes predicting a subsequent action of a target user through an optimal path prediction model that is trained in a form of a graph representing a user action trajectory of a session unit and recommending a path of the predicted action as an optimal path. The recommendation includes identifying a user intent based on a previous action trajectory of the target user in a current session, and guiding a path corresponding to the user intent as one of the optimal paths.

Abstract: In one example described herein a system can receive, by a gate analysis service, a quantum assembly language (QASM) file. The QASM file can define a quantum algorithm that can include logic gates that can be executed on a quantum computer system. The system can access, by the gate analysis service, a data repository that can include an estimated amount of quantum decoherence associated with each logic gate of a plurality of logic gates that includes the logic gates. The system can determine, by the gate analysis service, a prediction of an amount of quantum decoherence associated with executing at least one logic gate of the logic gates on the quantum computer system. Additionally, the system can adjust, by the gate analysis service, the QASM file to modify the prediction associated with executing the logic gates on the quantum computer system.

Abstract: A resource scheduling method and system, and a computer-readable storage medium are provided. The method comprises: a control instruction inputted by a user (S101) is acquired; on the basis of the control instruction, an intended device (S102) is determined; in the case where the intended device is not able to be started, with regard to each function of the intended device, a constructed knowledge graph is queried for alternative devices associated with the function, the knowledge graph having a plurality of functions stored therein as well as devices associated with the plurality of functions (S103, S105); and a device to be started is determined from all the alternative devices and starting the determined device (S106). In the resource scheduling method, in the case where an intended device specified by a user is not able to be started, alternative devices are appropriately determined.

Abstract: Methods, systems and apparatus for estimating an expectation value of a quantum mechanical observable. In one aspect, a method includes identifying a first operator associated with the observable, wherein the first operator comprises a linear combination of terms. One or more constraints on expectation values of one or more of the terms in the linear combination are determined. A second operator is defined, wherein the second operator comprises a combination of the first operator and one or more of the determined constraints. The expectation value of the quantum mechanical observable is estimated using the second operator.

Abstract: A non-transitory computer readable medium includes a set of instructions that in operation cause a processor to determine at least one modelled parameter of a feature of interest in petroleum exploration. The instructions also cause a processor to assign a likelihood value to each modelled parameter of the at least one modelled parameter and to generate an initial chance map for each modelled parameter of the at least one modelled parameter. Further, the instructions cause a processor to assign a weighting factor for each modelled parameter of the at least one modelled parameter, and to combine the initial chance maps using the weighting factor for each modelled parameter of the at least one modelled parameter to generate a first simulation chance map.

Abstract: A database node selects one of a number of database mirror hosts to perform a database read query. In theory, mirror hosts store redundant database entries. In practice, some mirror hosts fail to receive or retain some entries, e.g. if a mirror host was down for maintenance, a mirror host was overburdened, etc. The health of a mirror host, quantified as a health score, represents a likelihood that an insert statement was received and retained. Health scores are generated based on health metrics—data periodically retrieved from mirror hosts used to infer how well the database was operating, e.g. a count of entries that were inserted, a count of errors encountered while the entries were inserted, etc. If the database read query specifies entries inserted during a specific time period, the mirror host may be selected based on the health scores representing that time period.

Abstract: Various embodiments of the present disclosure provide computer interpretation techniques for implementing a query resolution process to improve upon traditional search resolutions within a search domain. The techniques may include generating a plurality of interaction embeddings for a plurality of textual descriptions corresponding to a plurality of interaction codes identified within an interaction dataset and a plurality of taxonomy embeddings for a plurality of taxonomy categories identified within a taxonomy dataset. The techniques may include generating similarity scores for a plurality of description-category pairs based on a comparison between the plurality of interaction embeddings and the plurality of taxonomy embeddings.

Abstract: A method and an apparatus for a messaging service are disclosed. A method of operating a terminal on which an application for a messaging service is installed includes detecting a keyword from a text input through a message input window, cumulatively displaying the detected keyword on a suggest area, setting information about a link to a selected keyword in the text, and separately displaying the selected keyword from a non-selected keyword on the suggest area.

Abstract: Methods and systems for fault identification and mitigation in an engine system. A state observer obtains current state information from the engine system, and a feature calculator uses data obtained from the state observer to calculate one or more feature indicators, which are monitored by a health estimator for the occurrence of a change using one or more change probability models. When the health estimator identifies a change, a fault isolator determines a component of the engine system that is subject to fault or health deterioration.

Abstract: A controller is provided for building equipment including a plurality of devices that operate in parallel to affect an environmental condition of a building. The controller includes one or more processing circuits including one or more processors and memory. The memory store instructions that, when executed by the one or more processors, cause the one or more processors to perform operations. The operations include lowering an upper bound or raising a lower bound of one or more constraints based on a minimum off schedule or a minimum on schedule for the building equipment, performing an optimization of an objective function subject to the one or more constraints to generate control decisions for the building equipment, and operating the building equipment in accordance with the control decisions to affect the environmental condition of the building.

Abstract: Methods, systems, and devices for wind turbine control based on reinforcement learning are disclosed. The method comprises receiving data indicative of a current environmental state of the wind turbine, determining one or more controlling actions of the wind turbine based on the current environmental state of the wind turbine and a reinforcement learning algorithm, and applying the determined one or more controlling actions to the wind turbine.

Abstract: Mechanisms are provided for performing an automated machine learning (AutoML) operation to configure parameters of a machine learning model. AutoML logic is configured based on an initial parameter sampling configuration for sampling values of parameter(s) of the machine learning (ML) model. An initial AutoML process is executed on the ML model based on a dataset utilizing the initially configured AutoML logic, to generate at least one learned value for the parameter(s) of the ML model. The dataset is analyzed to extract a set of dataset characteristics that define properties of a format and/or a content of the dataset which are stored in association with the at least one learned value as part of a training dataset. A ML prediction model is trained based on the training dataset to predict, for new datasets, corresponding new sampling configuration information based on characteristics of the new datasets.

Abstract: A method for mining an entity relationship, apparatus, electronic device, and a storage medium are provided. An implementation of the method may include: searching, in a query statement set, for a query statement satisfying a preset condition corresponding to a target entity word; in response to at least one query statement satisfying the preset condition corresponding to the target entity word being found in the query statement set, mining an associated entity word corresponding to the target entity word based on the found at least one query statement satisfying the preset condition corresponding to the target entity word.

Abstract: A method and system architecture for designing a compressive sensing matrix for machine learning includes receiving an image associated with a classification task and; generating a sensing matrix. The sensing matrix includes an array of nonzero elements of the image. A prism array of prism elements is in communication with the sensing matrix. A row of values corresponding with an input angle of the prism array is mapped to a respective column corresponding with a detector. Then the detector detects light refracted at an output angle dictated by the physical shape of the prism element. A physical model of the detector is fabricated and generates a compressed representation of the image. A machine learning classification algorithm is applied to the compressed representation of the image and generates an optimized non-invertible final determination of the image.