Abstract: A novel unified neural network framework, the dynamic memory network, is disclosed. This unified framework reduces every task in natural language processing to a question answering problem over an input sequence. Inputs and questions are used to create and connect deep memory sequences. Answers are then generated based on dynamically retrieved memories.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for reinforcement learning using agent curricula. One of the methods includes maintaining data specifying plurality of candidate agent policy neural networks; initializing mixing data that assigns a respective weight to each of the candidate agent policy neural networks; training the candidate agent policy neural networks using a reinforcement learning technique to generate combined action selection policies that result in improved performance on a reinforcement learning task; and during the training, repeatedly adjusting the weights in the mixing data to favor higher-performing candidate agent policy neural networks.

Abstract: An apparatus, a computer readable medium, and a learning method for learning a model corresponding to time-series input data, including acquiring the time-series input data, which is a time series of input data including a plurality of input values, propagating, to a plurality of nodes in a model, each of a plurality of propagation values obtained by weighting each input value at a plurality of time points before one time point according to passage of time points, in association with the plurality of input values at the one time point, calculating a node value of a first node among the plurality of nodes by using each propagated value propagated to the first node, and updating a weight parameter used to calculate each propagation value propagated to the first node, by using a corresponding input value and a calculated error of the node value at the one time point.

Abstract: A method includes defining model attributes of an organizational machine model that organizes feedback data from one or more sources of the feedback data into plural different topic groups based on similarities in concepts expressed in the feedback data. The model attributes represent criteria for establishment of the organizational machine model and include a topic model number that defines how many of the different topic groups are to be created by the organizational machine model and used to organize the feedback data into, a hyperparameter optimization alpha value that defines how likely a feedback datum in the feedback data is to be included in a single topic group of the different topic groupings or multiple topic groups of the different topic groupings, and a hyperparameter optimization beta value that defines how broadly each of the different topic groups are defined relative to the feedback data.

Abstract: A method, including: performing, with a computer, within-seismic-attribute clustering for each of a plurality of seismic attribute datasets for N different attributes, N being greater than or equal to two; identifying an anchor attribute and N?1 subordinate attributes from the N different attributes; linking, with a computer, objects within the seismic attribute data sets corresponding to the N?1 subordinate attributes to related objects within the seismic attribute data set corresponding to the anchor attribute; and identifying, with a computer, cross-attribute clusters, wherein the objects of any subordinate attribute that are linked to a same object of the anchor attribute are part of a single cross-attribute cluster.

Abstract: Techniques proactively deploy analytics to a computerized edge device. The techniques involve receiving data from the edge device. The data is conveyed through the edge device from a set of sensors disposed at a particular location. The techniques further involve performing analytics on the data to identify a set of edge device rules that defines a set of actions for the edge device to carry out under a set of predefined conditions potentially sensed by the set of sensors. The techniques further involve providing a command to the edge device. The command (i) includes the set of edge device rules and (ii) directs the edge device to, at a future time, start operating according to the set of edge device rules to protect against unsuccessful deployment of the command to the edge device due to subsequent delayed communication between the processing circuitry and the edge device.

Abstract: An information processing system, which includes a control system and an artificial neural network, is disclosed. The artificial neural network includes a group of neurons and a group of synapses, which includes a first portion and a second portion. The control system selects one of a group of operating modes. The group of neurons processes information. The group of synapses provide connectivity to each of the group of neurons. During a first operating mode of the group of operating modes, the first portion of the group of synapses is enabled and the second portion of the group of synapses is enabled. During a second operating mode of the group of operating modes, the first portion of the group of synapses is enabled and the second portion of the group of synapses is disabled.

Abstract: A learning method for acquiring at least one personalized reward function, used for performing a Reinforcement Learning (RL) algorithm, corresponding to a personalized optimal policy for a subject driver is provided. And the method includes steps of: (a) a learning device performing a process of instructing an adjustment reward network to generate first adjustment rewards, by referring to the information on actual actions and actual circumstance vectors in driving trajectories, a process of instructing a common reward module to generate first common rewards by referring to the actual actions and the actual circumstance vectors, and a process of instructing an estimation network to generate actual prospective values by referring to the actual circumstance vectors; and (b) the learning device instructing a first loss layer to generate an adjustment reward and to perform backpropagation to learn parameters of the adjustment reward network.

Abstract: A machine learning device provided in a controller for controlling a wire electrical discharge machine uses state variables (including data relating to a correction amount, a machining path, machining conditions, and a machining environment) observed by a state observation unit and determination data acquired by a determination data acquisition unit to machine-learn a correction for a machining path. Using the learning result, the machining path can be corrected automatically and accurately on the basis of a partial machining path, the machining conditions and the machining environment of the machining performed by the wire electrical discharge machine.

Abstract: A method, system, and computer program product for weight adjusted composite model for forecasting in anomalous environments are provided in the illustrative embodiments. A base forecasting model and a second forecasting model are combined to form a composite model, the base forecasting model configured to forecast an event in a time series, the second forecasting model configured to represent an anomalous portion of data in the time series. A mixing algorithm is combined with the composite model to adjust a set of weights associated with the composite model. Upon identifying a future period in which the event is to be forecasted, using the mixing algorithm, a subset of the set of weights is adjusted to from a weight adjusted composite model. The weight adjusted composite model is executed to forecast the event in the future period.

Abstract: A computer-implemented method includes recording, with a three-dimensional camera, one or more demonstrations of a user performing one or more reaching tasks. Training data is computed to describe the one or more demonstrations. One or more weights of a neural network are learned based on the training data, where the neural network is configured to estimate a goal location of the one or more reaching tasks. A partial trajectory of a new reaching task is recorded. An estimated goal location is computed, by a computer processor, by applying the neural network to the partial trajectory of the new reaching task.

Abstract: The technology disclosed provides a so-called “joint many-task neural network model” to solve a variety of increasingly complex natural language processing (NLP) tasks using growing depth of layers in a single end-to-end model. The model is successively trained by considering linguistic hierarchies, directly connecting word representations to all model layers, explicitly using predictions in lower tasks, and applying a so-called “successive regularization” technique to prevent catastrophic forgetting. Three examples of lower level model layers are part-of-speech (POS) tagging layer, chunking layer, and dependency parsing layer. Two examples of higher level model layers are semantic relatedness layer and textual entailment layer. The model achieves the state-of-the-art results on chunking, dependency parsing, semantic relatedness and textual entailment.

Abstract: Examples relate to providing hierarchical classifiers. In some examples, a superclass classifier of a hierarchy of classifiers is trained with a first type of prediction threshold, where the superclass classifier classifies data into one of a number of subclasses. At this stage, a subclass classifier is trained with a second type of prediction threshold, where the subclass classifier classifies the data into one of a number of classes. The first type of prediction threshold of the superclass classifier and the second type of prediction threshold of the subclass classifier are alternatively applied to classify data segments.

Abstract: A system for determining data requirements to generate machine-learning models. The system may include one or more processors and one or more storage devices storing instructions. When executed, the instructions may configure the one or more processors to perform operations including: receiving a sample dataset, generating a plurality of data categories based on the sample dataset; generating a plurality of primary models of different model types using data from the corresponding one of the data categories as training data; generating a sequence of secondary models by training the corresponding one of the primary models with progressively less training data; identifying minimum viable models in the sequences of secondary models; determining a number of samples required for the minimum viable models; and generating entries in the database associating: model types; corresponding data categories; and corresponding numbers of samples in the training data used for the minimum viable models.

Abstract: A system and method provide a sequence learning model. The method for training the sequence learning model comprises retrieving input sequence data. The input sequence data includes one or more input time sequences. The method also encodes the input sequence data into output symbol data using a sequence learning model. The output symbol data includes one or more symbolic representations. The method decodes, based on a neural network, the output symbol data to decoded sequence data, where the decoded sequence data includes one or more decoded time sequences that are to match the one or more input time sequences in the input sequence data. The method further compares the decoded sequence data with the input sequence data and updates the sequence learning model based on the comparison.

Abstract: Evolution and coevolution of neural networks via multitask learning is described. The foundation is (1) the original soft ordering, which uses a fixed architecture for the modules and a fixed routing (i.e. network topology) that is shared among all tasks. This architecture is then extended in two ways with CoDeepNEAT: (2) by coevolving the module architectures (CM), and (3) by coevolving both the module architectures and a single shared routing for all tasks using (CMSR). An alternative evolutionary process (4) keeps the module architecture fixed, but evolves a separate routing for each task during training (CTR). Finally, approaches (2) and (4) are combined into (5), where both modules and task routing are coevolved (CMTR).

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for determining neural network architectures. One of the methods includes generating, using a controller neural network, a batch of output sequences, each output sequence in the batch defining a respective architecture of a child neural network that is configured to perform a particular neural network task; for each output sequence in the batch: training a respective instance of the child neural network having the architecture defined by the output sequence; evaluating a performance of the trained instance of the child neural network on the particular neural network task to determine a performance metric for the trained instance of the child neural network on the particular neural network task; and using the performance metrics for the trained instances of the child neural network to adjust the current values of the controller parameters of the controller neural network.

Abstract: A learning apparatus includes at least one memory and at least one circuit. The circuit (a) obtains a first neural network that has learned by using source learning data and obtains target learning data, the target learning data including a plurality of first data items each of which is given a first label and a plurality of second data items each of which is given a second label, (b) obtains a plurality of first output vectors by inputting the plurality of first data items to a second neural network and obtains a plurality of second output vectors by inputting the plurality of second data items to the second neural network, and (c) generates a first relation vector corresponding to the first label by using the plurality of first output vectors and generates a second relation vector corresponding to the second label by using the plurality of second output vectors.

Abstract: A method for improving a profile analysis of an interpretive framework stored in a memory may include producing and displaying visual stimuli on a computerized device to test visual and visual motor responses of an individual subject in response to the displayed visual stimuli. The method may also include classifying and categorizing digitally measured visual and visual motor responses of the individual subject to the displayed visual stimuli. The method may further include continually modifying parameters of the profile analysis of the interpretive framework corresponding to at least one condition based at least in part on an item analysis corresponding to a pattern of performance determined during the classifying and categorizing of the digitally measured visual and visual motor responses of the individual subject.

Abstract: A programming method for an artificial neuron network having synapses, each including a single resistive random-access memory having first and second electrodes on either side of an active zone, the method including determining a number N of conductance intervals, where N?3; for each memory: choosing a conductance interval from amongst the N intervals; a step i) for application of a voltage pulse of a first type between the first and second electrodes, and for reading the conductance value of the memory; if the conductance value does not belong to the previously chosen conductance interval, a sub-step ii) for application of a voltage pulse of a second type between the first and second electrodes, and for reading the conductance value; if the conductance value does not belong to the chosen conductance interval, a step according to which step i) is reiterated, with steps i) and ii) being repeated until the conductance value belongs to the interval.