Abstract: A technique for building a modified entry trajectory profile for an aircraft to join a holding pattern at an entry waypoint despite the airspace available to build the entry trajectory being limited. In one aspect, instructions are received for an aircraft to join a holding pattern at an entry waypoint according to an entry trajectory profile. An airspace around the holding pattern is divided into sectors using the entry waypoint as a reference point. A discontinuity in the entry trajectory profile can be identified. In response to identifying the discontinuity in the entry trajectory profile, a modified entry trajectory profile can be built by i) determining the sector from which the aircraft would approach the entry waypoint with a current track angle of the aircraft moved to intersect the entry waypoint and with the aircraft pointing at the entry waypoint; and ii) generating the modified entry trajectory profile.

Abstract: A computer-implemented method of training an auxiliary machine learning model to predict a set of new parameters of a primary machine learning model, wherein the primary model is configured to transform from an observed subset of a set of real-world features to a predicted version of the set of real-world features.

Abstract: A computer-implemented method of training an auxiliary machine learning model to predict a set of new parameters of a primary machine learning model, wherein the primary model is configured to transform from an observed subset of a set of real-world features to a predicted version of the set of real-world features.

Abstract: A self-reconfigurable controller for a robot, including: input/output (I/O) interfaces to enable communication with I/O peripheral devices coupled to the robot; and processing circuitry that is operable to: register the I/O peripheral devices and associated functionalities; receive a command for the robot to perform a task; conduct a self-awareness check to correlate functionalities to perform the task with functionalities of the I/O peripheral devices; and generate, based on a net of deep learning (DL) models and a result of the correlation, a target deep learning controller (TDLC) model to control the robot to perform the task.

Abstract: An equipment controller can include a processor; memory accessible to the processor; and processor-executable instructions stored in the memory to instruct the equipment controller to: instantiate an edge application and an edge framework, where the edge framework includes a framework engine; receive sensor data; process the sensor data via the edge application to issue a call to the edge framework; responsive to the call, implement the framework engine to generate a result; and based at least in part on the result, issue an equipment control signal.

Abstract: A method of operating a neural network, comprising: at each input node of an input layer, weighting a respective input element received by that node by applying a first class of probability distribution, thereby generating a respective set of output parameters describing an output probability distribution; and from each input node, outputting the respective set of output parameters to one or more nodes in a next, hidden layer of the network, thereby propagating the respective set of output parameters through the hidden layers to an output layer; the propagating comprising, at one or more nodes of at least one hidden layer, combining the sets of input parameters and weighting the combination by applying a second class of probability distribution, thereby generating a respective set of output parameters describing an output probability distribution, wherein the first class of probability distribution is more sparsity inducing than the second class of probability distribution.

Abstract: A method of training a model comprising a generative network mapping a latent vector to a feature vector, wherein weights in the generative network are modelled as probabilistic distributions. The method comprises: a) obtaining one or more observed data points, each comprising an incomplete observation of the features in the feature vector; b) training the model based on the observed data points to learn values of the weights of the generative network which map the latent vector to the feature vector; c) from amongst a plurality of potential next features to observe, searching for a target feature of the feature vector which maximizes a measure of expected reduction in uncertainty in a distribution of said weights of the generative network given the observed data points so far; and d) outputting a request to collect a target data point comprising at least the target feature.

Abstract: A method of training a model comprising a generative network mapping a latent vector to a feature vector, wherein weights in the generative network are modelled as probabilistic distributions. The method comprises: a) obtaining one or more observed data points, each comprising an incomplete observation of the features in the feature vector; b) training the model based on the observed data points to learn values of the weights of the generative network which map the latent vector to the feature vector; c) from amongst a plurality of potential next features to observe, searching for a target feature of the feature vector which maximizes a measure of expected reduction in uncertainty in a distribution of said weights of the generative network given the observed data points so far; and d) outputting a request to collect a target data point comprising at least the target feature.

Abstract: A computer-implemented method comprising: outputting questions to a user via one or more user devices, and receiving back responses to some of the questions from the user via one or more user devices; over time, controlling the outputting of the questions so as to output the questions under circumstances of different values for each of one or more items of metadata, wherein the one or more items of metadata comprise one or more physical conditions of the user; monitoring whether or not the user responds when the question is output with the different metadata values; training the machine learning algorithm to learn a value of each of the items of metadata which optimizes a reward function, and based thereon selecting a circumstance when the user is exhibiting a particular physical condition to output subsequent questions.

Abstract: A method of monitoring execution of computer instructions includes receiving data items representing real-time measurements of side-channel information emanating from execution of computer instructions, each data item forming a value of a corresponding dimension of a side-channel information vector, receiving, for two or more of the dimensions of the side-channel vector, classifiers that assign the corresponding values of a side-channel vector to classes, and classifying the data items in accordance with the received classifiers, wherein an orthogonal distance of the data item from the classifier indicates a confidence value of the classification, generating a combined a confidence value for the side-channel information vector a, and outputting a signal if a confidence value indicates affiliation to a selected one of the two classes with a predetermined probability. The method conducts a self-test by generating a combined confidence value based to ensure correct outputting of the confidence value.

Abstract: A computer-implemented method comprising: outputting questions to a user via one or more user devices, and receiving back responses to some of the questions from the user via one or more user devices; over time, controlling the outputting of the questions so as to output the questions under circumstances of different values for each of one or more items of metadata, wherein the one or more items of metadata comprise at least a time and/or a location at which a question was output to the user via the one or more user devices; monitoring whether or not the user responds when the question is output with the different metadata values; training the machine learning algorithm to learn a value of each of the items of metadata which optimizes a reward function, and based thereon selecting a time and/or location at which to output subsequent questions.

Abstract: A mobile device input device connected to a mobile device, the mobile device input device including a main body to be worn by a user, a control unit running a program thereon and disposed within at least a portion of the main body to at least one operation of the mobile device, and a laser sensor disposed on at least a portion of the main body to emit a laser beam away from a surface of the main body to track a position of the main body.

Abstract: A base slat pry stopper, such as a base slat locking bracket, for locking a rolling shutter curtain covering an opening of a structure is disclosed. The base slat pry stopper may include a bracket base plate mounted to an opening bottom wall on an interior side of the shutter curtain, and a base slat engagement portion extending upward from the bracket base plate and above a top surface of the opening bottom wall. When the shutter curtain is unrolled to cover the opening, the base slat engagement portion engages the base slat to prevent the base slat from deflecting upward away from the opening bottom wall when a force applied from an exterior side of the shutter curtain causes the base slat to deflect toward the interior side of the structure and the shutter curtain. The base slat pry stopper may be permanently or removably installed.

Abstract: Embodiments relate to a system, comprising, an electromechanical platform for positioning and orienting a collectible to capture a plurality of images by an image capturing device of at least a first side and a second side of the collectible, a computer comprising at least one processor comprising computer-executable instructions stored on one or more computer-readable media, wherein the computer is operable to receive the plurality of images of the collectible, at least one processing routine comprising an image processing algorithm for a condition assessment of the collectible applied to at least one image from the plurality of images by at least one processor to obtain a raw data of the condition of the collectible, a device for encapsulating the collectible in a tamper proof casing and stamping at least one or more labels on the tamper proof casing, wherein the stamping of the at least one or more labels is at a specified location on the tamper proof casing; and wherein the system is operable to be fully

Abstract: A computer-implemented method of training an auxiliary machine learning model to predict a set of new parameters of a primary machine learning model, wherein the primary model is configured to transform from an observed subset of a set of real-world features to a predicted version of the set of real-world features.

Abstract: Apparatus and methods are described in which a namespace is protected against misleading registration of names within the namespace. A list of canonically expressed text strings is maintained. A concordancer is defined, which concordancer may be updated from time to time as, for example, characters are added to a permitted character set for the namespace. When a first user attempts to register a proposed name within the namespace, the proposed name is subjected to an attempted match to each of the outputs of the concordancer with respect to each of the canonically expressed text strings on the list. In the event of a match, the attempted registration is not permitted to proceed. The protection may be implemented simultaneously across multiple namespaces, each having its own respective concordancer.

Abstract: A base slat pry stopper, such as a base slat locking bracket, for locking a rolling shutter curtain covering an opening of a structure is disclosed. The base slat pry stopper may include a bracket base plate mounted to an opening bottom wall on an interior side of the shutter curtain, and a base slat engagement portion extending upward from the bracket base plate and above a top surface of the opening bottom wall. When the shutter curtain is unrolled to cover the opening, the base slat engagement portion engages the base slat to prevent the base slat from deflecting upward away from the opening bottom wall when a force applied from an exterior side of the shutter curtain causes the base slat to deflect toward the interior side of the structure and the shutter curtain. The base slat pry stopper may be permanently or removably installed.

Abstract: In a first stage, training each of a plurality of first variational auto encoders, VAEs, each comprising: a respective first encoder arranged to encode a respective subset of one or more features of a feature space into a respective first latent representation, and a respective first decoder arranged to decode from the respective latent representation back to a decoded version of the respective subset of the feature space, wherein different subsets comprise features of different types of data. In a second stage following the first stage, training a second VAE comprising: a second encoder arranged to encode a plurality of inputs into a second latent representation, and a second decoder arranged to decode the second latent representation into decoded versions of the first latent representations, wherein each of the plurality of inputs comprises a combination of a different respective one of feature subsets with the respective first latent representation.

Abstract: A blockchain arrangement configured to simultaneously distribute at least one public transaction and/or a restricted transaction, wherein the disposition includes a plurality of participating nodes and a plurality of validator nodes connected by a telecommunications network; wherein a proponent participant node is configured to send to all the validator nodes the contents of a tx information transaction record, together with the identifier of a particular preconfigured privacy group; to provide the capabilities of anonymity and privacy to the blockchain system that distributes blocks of information transaction records; such that at least one recipient participating node, connected to a blockchain network, is able to decrypt, read, and execute the information transaction record blocks encrypted by the validator or mining nodes of the blockchain network.

Abstract: A method of training a model comprising a generative network mapping a latent vector to a feature vector, wherein weights in the generative network are modelled as probabilistic distributions. The method comprises: a) obtaining one or more observed data points, each comprising an incomplete observation of the features in the feature vector; b) training the model based on the observed data points to learn values of the weights of the generative network which map the latent vector to the feature vector; c) from amongst a plurality of potential next features to observe, searching for a target feature of the feature vector which maximizes a measure of expected reduction in uncertainty in a distribution of said weights of the generative network given the observed data points so far; and d) outputting a request to collect a target data point comprising at least the target feature.