Abstract: Embodiments of the present invention provide a method for detecting a temporal change of name associated with performance data. The method comprises receiving at least one candidate name replacement pair comprising a pair of names. The method further comprises, in a training stage, for each known name replacement pair included in the performance data, determining a window of time covering a most recent appearance of a first name of the known name replacement pair. The window of time is determined based on quantitative features of a time series model comprising performance data for the first name and a second name of the known name replacement pair. The method further comprises, in the training stage, training a machine learning classifier based on quantitative features computed using a portion of the performance data for the first name and the second name, where the portion is within the window of time determined.

Abstract: A method for characterizing a user associated with a vehicle including collecting a movement dataset sampled at least at one of a location sensor and a motion sensor associated with the vehicle, during a time period associated with movement of the vehicle; extracting a set of movement features associated with movement of at least one of the user and the vehicle during the time period; and determining one or more user characteristics describing the user based on the set of movement features, wherein the one or more user characteristics include a classification of the user as at least one of a passenger and a driver for the time period associated with movement of the vehicle.

Abstract: The present disclosure relates generally to mechanisms for the estimation of location privacy risk, comprising: building one or more trajectory models from auxiliary information (e.g., one or more maps, one or more routes); capturing common behavioral patterns (e.g., shortest route(s),/fastest route(s)); identifying, given unlinked trajectories for a plurality of users, most likely linkages using the trajectory model(s); eliminating one or more unlikely linkages based on deviation from the shortest route(s) and/or the fastest route(s); measuring privacy as the percentage of linkages correctly identified; and outputting the measured privacy.

Abstract: A method includes determining a trainable model to provide to a trainer, the trainable model determined based on modification of one or more models of a plurality of models. The plurality of models is generated based on a genetic algorithm and corresponds to a first epoch of the genetic algorithm. Each of the plurality of models includes data representative of a neural network. The method also includes providing the trainable model to the trainer. The method further includes adding a trained model, output by the trainer based on the trainable model, as input to a second epoch of the genetic algorithm, the second epoch subsequent to the first epoch.

Abstract: Apparatus and method for a scalable, free running neuromorphic processor. For example, one embodiment of a neuromorphic processing apparatus comprises: a plurality of neurons; an interconnection network to communicatively couple at least a subset of the plurality of neurons; a spike controller to stochastically generate a trigger signal, the trigger signal to cause a selected neuron to perform a thresholding operation to determine whether to issue a spike signal.

Abstract: In some example embodiments, a system and method are provided for graph pattern analysis. In example embodiments, pattern data of a primary network that includes data relating to relationships between entities are received. A reference matrix comprising a plurality of secondary network determined based on the pattern data of the primary network is generated. A graphical display of the primary network and the reference matrix are displayed on a user interface. A selection of a secondary network from the plurality of secondary networks of the reference matrix is received. The selected secondary network has similar matching characteristics with at least a portion of the primary network. In response to the selection, the primary network is classified as a classification type related to the selected secondary network.

Abstract: Anomaly scores for respective message types in computer log data and confidence intervals for respective anomaly scores are calculated based on a number of appearances of respective message types in a plurality of models generated from a historical set of computer log data. Respective models of the plurality of models can have at least a portion of the historical set of computer log data excluded from the respective models. Respective anomaly scores and respective confidence intervals can be applied to a new set of log data to identify and troubleshoot unusual log data events.

Abstract: Embodiments relate to systems, devices, and computing-implemented methods for providing a response system for a chatbot program and/or virtual assistant that can perform operations including receiving user input that includes an identifier, determining a current state based on the identifier, determining an intent engine stack based on the current state, processing the user input using an intent engine in the intent engine stack to obtain an intent and a variable, performing an action based on the current state, the intent, and the variable, and transitioning to a next state based on the action, the current state, the intent, and the variable.

Abstract: A CMOS-based resistive processing unit (RPU) and method for a neural network. The RPU includes a capacitor device configured to store a charge representing a weight value associated with a neural network circuit operation. A current source Field Effect Transistor (FET) device is operatively connected to the capacitor device for increasing a charge on the capacitor. A current sink FET device operatively connected to the capacitor device is configured for decreasing the stored capacitor charge. An analog weight update circuit receives one or more update signals generated in conjunction with the neural network circuit operation, the analog weight update circuit controlling the current source FET device and the current sink FET device to provide either a determined amount of current to increase the stored charge on the capacitor device, or sink a determined amount of current to decrease the stored charge on the capacitor device.

Abstract: A CMOS-based resistive processing unit (RPU) for a neural network. The RPU includes a capacitor device configured to store a charge representing a weight value associated with a neural network circuit operation. A current source Field Effect Transistor (FET) device is operatively connected to the capacitor device for increasing a charge on the capacitor. A current sink FET device operatively connected to the capacitor device is configured for decreasing the stored capacitor charge. An analog weight update circuit receives one or more update signals generated in conjunction with the neural network circuit operation, the analog weight update circuit controlling the current source FET device and the current sink FET device to provide either a determined amount of current to increase the stored charge on the capacitor device, or sink a determined amount of current to decrease the stored charge on the capacitor device.

Abstract: Systems, methods, and apparatus, including computer programs encoded on a computer storage medium, related to associative long short-term memory (LSTM) neural network layers configured to maintain N copies of an internal state for the associative LSTM layer, N being an integer greater than one. In one aspect, a system includes a recurrent neural network including an associative LSTM layer, wherein the associative LSTM layer is configured to, for each time step, receive a layer input, update each of the N copies of the internal state using the layer input for the time step and a layer output generated by the associative LSTM layer for a preceding time step, and generate a layer output for the time step using the N updated copies of the internal state.

Abstract: Systems, methods, devices, and other techniques are disclosed for using an augmented neural network system to generate a sequence of outputs from a sequence of inputs. An augmented neural network system can include a controller neural network, a hierarchical external memory, and a memory access subsystem. The controller neural network receives a neural network input at each of a series of time steps processes the neural network input to generate a memory key for the time step. The external memory includes a set of memory nodes arranged as a binary tree. To provide an interface between the controller neural network and the external memory, the system includes a memory access subsystem that is configured to, for each of the series of time steps, perform one or more operations to generate a respective output for the time step. The capacity of the neural network system to account for long-range dependencies in input sequences may be extended.

Abstract: A capacity-analysis tool (CAT) provides a model framework for creating a model of a capacity-planning-target (CPT) system, e.g., a data center. The tool includes a model framework that, in turn, includes a closed-system template for creating CSMs, i.e., models of capacity-limited systems. A user uses the CAT to create CPT models using the CSMs as building blocks. A machine-learning engine is used to train the CPT model, converting parameter time-series data to functions of time. The trained CPT models are then used to make capacity-planning estimates, e.g., time remaining on a system before usage matches capacity. The CAT makes it easy to extend a model, e.g., by adding new dimensions (new factors of interest) in the form of new CSMs to which the new dimensions have been assigned.

Abstract: A system and method for controlling a nodal network. The method includes estimating an effect on the objective caused by the existence or non-existence of a direct connection between a pair of nodes and changing a structure of the nodal network based at least in part on the estimate of the effect. A nodal network includes a strict partially ordered set, a weighted directed acyclic graph, an artificial neural network, and/or a layered feed-forward neural network.

Abstract: A method, non-transitory computer readable medium and apparatus for predicting mortality of a current patient are disclosed. For example, the method includes receiving data associated with a plurality of different patients with known mortality outcomes, wherein the data includes a subset of data for each one of a plurality of different measurement timepoints for each one of the plurality of different patients, calculating n number of classifiers, wherein n is equal to a number of the plurality of different measurement timepoints, receiving data associated with the current patient at an i-th measurement timepoint, predicting the current patient has a high mortality risk based on an output of the i-th classifier of the n number of classifiers and transmitting a signal to a health administration server to cause an alarm to be generated in response to the high mortality risk that is predicted.

Abstract: A machine learning device that performs reinforcement learning with respect to a servo control apparatus that controls target device having a motor, including: outputting action information including adjustment information of coefficients of a transfer function of a controller gain to a controller included in the servo control apparatus; acquiring, from the servo control apparatus, state information including a deviation between an actual operation of the target device and a command input to the controller, a phase of the motor, and the coefficients of the transfer function of the controller gain when the controller operates the target device based on the action information; outputting a value of a reward in the reinforcement learning based on the deviation included in the state information; and updating an action-value function based on the value of the reward, the state information, and the action information.

Abstract: A deep neural network architecture comprises a stack of strata in which each stratum has its individual input and an individual objective, in addition to being activated from the system input through lower strata in the stack and receiving back propagation training from the system objective back propagated through higher strata in the stack of strata. The individual objective for a stratum may comprise an individualized target objective designed to achieve diversity among the strata. Each stratum may have a stratum support subnetwork with various specialized subnetworks. These specialized subnetworks may comprise a linear subnetwork to facilitate communication across strata and various specialized subnetworks that help encode features in a more compact way, not only to facilitate communication across strata but also to increase interpretability for human users and to facilitate communication with other machine learning systems.

Abstract: In an example, one or more job postings, as well as corresponding confidential data values, are obtained from a social networking service. A first set of one or more features are extracted from the one or more job postings. The first set of one or more features and corresponding confidential data values are fed into a machine learning algorithm to train a confidential data value prediction model to output a predicted confidential data value for a candidate job posting. Then, the candidate job posting is obtained and a second set of one or more features are extracted from the candidate job posting. The extracted second set of one or more features is fed to the confidential data value prediction model, outputting the predicted confidential data value.

Abstract: T-shirt size recommendation for a first user based on crowd sourcing. A recommendation for t-shirt size may be made by defining a first image of a container based on historical data related to container instantiation; receiving, by the computer, from a first user, a request to deploy a first instantiation of a container corresponding to the first image, with the request including input data including information indicative of a planned context of the first instantiation; applying a plurality of machine logic based rules to the input data to determine a first a first recommended T-shirt size associated with the first image; and communicating, by the computer, the first recommended T-shirt size to the first user for use with the first instantiation of the container corresponding to the first image.

Abstract: A method and system are provided for updating synapse weight values in neuromorphic system with Spike Time Dependent Plasticity model. The method includes selectively performing, by a hardware-based synapse weight incrementer or decrementer, one of a synapse weight increment function or decrement function, each using a respective lookup table, to generate updated synapse weight values responsive to spike timing data. The method further includes storing the updated synapse weight values in a memory. The method additionally includes performing, by a hardware-based processor, a learning process to integrate the updated synapse weight values stored in the memory into the Spike Time Dependent Plasticity model neuromorphic system for improved neuromorphic simulation.