Abstract: Systems and techniques that facilitate computing touchpoint journey recommendations are provided. In various embodiments, an input component can receive a computing context of a client and a computing profile of a client. In various instances, the client can be engaged in a computing touchpoint journey. In various embodiments, a prediction component can predict, via a first machine learning classifier, a negative event likely to occur on the computing touchpoint journey. In various cases, the first machine learning classifier can receive as input the computing context and the computing profile and can generate as output the predicted negative event. In various embodiments, a decision component can recommend in real-time, via a second machine learning classifier, a computing touchpoint to which to transfer the client.

Abstract: A system, device, and method for implementing secure control over audio visual (AV) equipment connected to an AV gateway is disclosed. The solution implements secure and remote control over audio visual (AV) equipment included in an AV network by enabling a web browser running on a user device to utilize a web proxy shuttle to communicate control commands to an AV gateway that controls the AV equipment.

Abstract: Ways to provide information via dynamic user interfaces including Dynamic InfoGraphics (DIG) are described. DIG representations may include various elements, such as text, icons and other graphics, tags, links, etc. and may be implemented via various types of displays or operating environments. Users may include, for example, pilots, soldiers, commanders, drivers, game players, etc. Information provided via DIG may include, for instance, data received from various types of sensors, communications received from other parties or resources, calculated or otherwise determined information (e.g., predictions associated with artificial intelligence (AI) resources such as autonomous or semi-autonomous intelligent agents), and/or other appropriate information (e.g., user inputs). The received information may be applied to one or more DIG representations or templates. AI features may be used to predict future actions, paths, goals or targets, etc.

Abstract: A computing system, including a processor configured to, at development time, receive a machine learning model topology including a plurality of layers. The processor may be further configured to generate an internal representation graph of the machine learning model topology. The internal representation graph may include a plurality of internal representation layers. By performing one or more modifications to the internal representation graph, the processor may be further configured to generate an inferencer graph including a plurality of inferencer layer blocks. Each inferencer layer block may indicate an input buffer size, a logic function, and an output buffer size. At deployment time, the processor may be further configured to transmit, to a plurality of processing devices, instructions to implement the machine learning model topology with the respective input buffer sizes, logic functions, and output buffer sizes selected for the plurality of inferencer layer blocks of the inferencer graph.

Abstract: A building energy management includes building equipment, one or more data platform services, a timeseries database, and an energy management application. The building equipment operate to monitor and control a variable and provide raw data samples of a data point associated with the variable. The timeseries database stores a plurality of timeseries associated with the data point. The plurality of timeseries include a timeseries of the raw data samples and the one or more optimized data timeseries generated by the data platform services based on the raw data timeseries. The energy management application generates an ad hoc dashboard including a widget and associates the widget with the data point. The widget displays a graphical visualization of the plurality of timeseries associated with the data point and includes interactive user interface options for switching between the plurality of timeseries associated with the data point.

Abstract: The present disclosure includes methods and systems for generating digital predictive models by progressively sampling a repository of data samples. In particular, one or more embodiments of the disclosed systems and methods identify initial attributes for predicting a target attribute and utilize the initial attributes to identify a coarse sample set. Moreover, the disclosed systems and methods can utilize the coarse sample set to identify focused attributes pertinent to predicting the target attribute. Utilizing the focused attributes, the disclosed systems and methods can identify refined data samples and utilize the refined data samples to identify final attributes and generate a digital predictive model.

Abstract: Data-parallel ensemble training using gradient boosted trees includes training an ensemble of trees. The training includes splitting a training dataset into several data portions. Each data portion is assigned to each thread group from a set of thread groups. The training further includes executing a stage, in which each thread group, in parallel, trains a respective ensemble of decision trees. Executing the stage includes performing, by each thread group, in parallel, machine learning operations for the respective ensemble of decision trees using the data portion assigned to each thread group. Further, each thread group validates, in parallel, the respective ensemble of decision trees using a data portion assigned to another thread group. Execution of the stage is repeated until a predetermined threshold is satisfied. Further, a prediction is inferenced using the ensemble of decision trees that is formed using the respective ensemble of trees from each of the thread groups.

Abstract: The disclosed techniques include obtaining sensor data from one or more sensors configured to detect movement related to a user of the one or more sensors, and determining, based on the sensor data, of an occurrence of a gesture-based physical behavior event of the user. The techniques further cause adjustment of a medication dosage for delivery to the user based on determining the occurrence of the gesture-based physical behavior event.

Abstract: A tri-state neural network circuit includes, in an intermediate layer: an input value Xi to which convolution is applied; a nonzero convolution operation circuit configured to receive a weight Wi and to perform a convolution operation thereof; a sum circuit configured to take a sum of operation values each subjected to the convolution operation and a bias W0; and an activating function circuit configured to convert a signal Y generated by taking the sum, using an activating function f(u). The nonzero convolution operation circuit is configured to skip a weight Wi having a zero weight and to perform a convolution operation based on a nonzero weight and an input value Xi corresponding thereto.

Abstract: A system and method for ensuring application content is displayed in a proper sized display is provided. An illustrative web-based application detects the display characteristics, e.g., size and/or aspect ratio, of its display window. If the display characteristics do not match a predefined requirement, the application ceases updating content and displays resizing notifications to alert the user to adjust the window. Once the window's display characteristics match the predefined criteria, the application resumes updating content in the window.

Abstract: Described embodiments provide systems and methods for defining a clip within a recording. A device may generate a recording of a session of a user via which a plurality of applications is accessible. The device may detect a transition of an application of the plurality of applications in a foreground of the session. The device may identify, from the recording, a frame corresponding to the transition in the foreground. The device may store an identification of the frame corresponding to the transition to define a clip of the user within the recording using the application in the session.

Abstract: A building energy management system includes building equipment, a data collector, an analytics service, a timeseries database, and an energy management application. The building equipment monitor and control one or more variables in the building energy management system and provide data samples of the one or more variables. The data collector collects the data samples from the building equipment and generates a data timeseries including a plurality of the data samples. The analytics service performs one or more analytics using the data timeseries and generates a results timeseries including a plurality of result samples indicating results of the analytics. The timeseries database stores the data timeseries and the results timeseries. The energy management application retrieves the data timeseries and the results timeseries from the timeseries database in response to a request for timeseries data associated with the one or more variables.

Abstract: Generally described, the presently disclosed technology utilizes managed Wi-Fi networks pre-installed throughout an MDU property to provide user-specific passphrases that can be used to access the single-SSID wireless network at the property and to provide a cloud portal that can enable convenient access to the functionalities (both by the resident and the manager) provided by the Wi-Fi controller and the Wi-Fi access points. By doing so, the Wi-Fi network management solutions described herein allow the users to experience the benefits of a shared Wi-Fi infrastructure, such as not having to set up and maintain their own Wi-Fi routers, while also allowing them to easily change their Wi-Fi settings from their connected devices.

Abstract: Provided herein are method, apparatus, and computer program products for facilitating capture and play back of interface event data. One exemplary method may comprise providing, as a function of at least profile data, environmental data, transaction data, relevance data, and implicit data derived therefrom, a dynamic application environment, displaying the dynamic application environment on an dynamic application interface, capturing interface event data, the interface event data being based on user interaction with the dynamic application interface, and comprising of one or more interface events and a particular time at which each of the one or more interface events occurred, providing data to the play back device configured for initializing a play back environment, such that the play back environment receives the interface event data and performs play back, and outputting the one or more interface events in a relative time to the particular time.

Abstract: Introduced here are teaching system for realistically simulating medical situations in such a manner to test individuals as to the adequacy of their learning. A teaching system may include a manikin representative of a patient, an image projection mechanism, a simulation mechanism, a headpiece, a pair of hand coverings, a controller, a performance review mechanism, a recording device, a scenario generation mechanism, or any combination thereof. The teaching system can improve upon conventional simulation in several respects. First, the script corresponding to a simulation session may be dynamically traversed to mimic the unpredictability of an actual medical situation. Second, the action(s) of an individual may be actively monitored to discover whether her behavior complied with the relevant protocol(s).

Abstract: The present disclosure describes a method, apparatus, and storage medium for generating network representation for a neural network. The method includes obtaining a source-side vector sequence corresponding to an input sequence.

Abstract: An acquisition unit acquires image data by performing screen capture of a screen operated by a user, at a predetermined interval. A detection unit detects the user operation. An estimation unit estimates an amount of delay time from a user operation time point to a time point at which the user operation is detected by the detection unit. The selection unit selects image data of the user operation time point out of the acquired image data.

Abstract: Methods and systems are disclosed for updating learned models. An embodiment comprises receiving a plurality of data sets representing sensed data from one or more devices and determining, using one or more local learned models, local parameters based on the received data sets. Another operation may comprise generating a combined data set by combining the plurality of data sets and, determining, using one or more local learned models, global parameters based on the combined data set. Another operation may comprise transmitting, to a remote system, the global parameters for determining updated global parameters using one or more global learned models based at least partially on the global parameters, and receiving, from the remote system, the updated global parameters. Another operation may comprise updating the one or more local learned models using both the local parameters and updated global parameters.

Abstract: An example method of identification of related event groups for a service monitoring system includes: receiving a sample set of events from a service monitoring system; choosing, based the sample set of events, a set of fieldnames for defining factors; generating a plurality of event group definitions, wherein each event group definition of the plurality of event group definitions comprises a plurality of factors, wherein each factor of the plurality of factors is represented by a respective fieldname-value pair of an event of the sample set of events, wherein a fieldname of the respective fieldname-value pair is selected from the set of fieldnames; and determining, based on the plurality of event group definitions, one or more event groups for a plurality of events.

Abstract: Apparatus and methods are disclosed for selection of characters from an electronic user interface. In one embodiment, a letter selection interface displays letters in a row, some of which are recordable by an input interpreter and others that are unrecordable. In one embodiment, unrecordable letters visually separate letters that are both (1) recordable, and (2) assigned to different but adjacent selection buttons. In response to received touch-screen activations, an interpreter records recordable letter selections and a display displays both recordable and unrecordable letter selections. The interpreter composes a sequence of the recorded letters. An algorithm identifies alternatives to the recorded sequence by editing one or more letters of the sequence, such that each edit changes a selected letter to an adjacent unrecordable letter. The algorithm outputs words that match the originally composed sequence or one of the identified alternatives.