Abstract: A neural network training apparatus is described which has a network of worker nodes each having a memory storing a subgraph of a neural network to be trained. The apparatus has a control node connected to the network of worker nodes. The control node is configured to send training data instances into the network to trigger parallelized message passing operations which implement a training algorithm which trains the neural network. At least some of the message passing operations asynchronously update parameters of individual subgraphs of the neural network at the individual worker nodes.

Abstract: Systems and methods for creating and/or using an artificial intelligence memory system that models human memory are provided. The AI memory system creates and/or uses a user centric memory graph. The user centric memory graph implicitly links memory elements of a user utilizing relationships created in space, time, and cognitive dimensions similar to how the human brain stores and recalls different memory elements. The user centric memory graph is used by searching and/or constraining the user centric memory graph based on a determined user context and/or a user query.

Abstract: The present disclosure discloses method and system for extracting relevant entities from a text corpus. The method comprises receiving, by the entity extraction computing device, a text corpus and an entity, determining at least one feature for each block of text from the text corpus, where the at least one feature corresponds to predefined one or more feature heads, calculating a score for each block of text from the text corpus based on training of the entity extraction system, determining a template from one or more templates based on the score, where the one or more templates are generated based on the training of the entity extraction system, and extracting at least one relevant entity from the text corpus, with respect to the entity, based on the template. The method and system disclosed in the present disclosure may be used to extract relevant entities across various domains by training the system.

Abstract: A synchrophasor measurement-based disturbance identification method is described considering different penetration levels of renewable energy. A differential Teager-Kaiser energy operator (dTKEO)-based algorithm is first utilized to improve multiple-disturbances detection accuracy. Then, feature extractions via the integrated additive angular margin (AAM) loss and the long short-term memory (LSTM) network is described. This enables one to deal with intra-class similarity and inter-class variance of disturbances when high penetration renewable energy occurs. With the extracted features, a multi-stage weighted summing (MSWS) loss-based criterion is described for adaptive data window determination and fast disturbance pre-classification. Finally, the re-identification model based on feature similarity is established to identify unknown disturbances, a challenge for existing machine learning algorithms.

Abstract: Computer-implemented, machine-learning systems and methods relate to a combination of neural networks. The systems and methods train the respective member networks both (i) to be diverse and yet (ii) according to a common, overall objective. Each member network is trained or retrained jointly with all the other member networks, including member networks that may not have been present in the ensemble when a member is first trained.

Abstract: In one aspect, this specification describes a recurrent neural network system implemented by one or more computers that is configured to process input sets to generate neural network outputs for each input set. The input set can be a collection of multiple inputs for which the recurrent neural network should generate the same neural network output regardless of the order in which the inputs are arranged in the collection. The recurrent neural network system can include a read neural network, a process neural network, and a write neural network. In another aspect, this specification describes a system implemented as computer programs on one or more computers in one or more locations that is configured to train a recurrent neural network that receives a neural network input and sequentially emits outputs to generate an output sequence for the neural network input.

Abstract: A set of vectors may be obtained. The vectors may be multi-dimensional vectors that are associated with and describe tokens from a first set of tokens from a corpus of sources. The description may be based in part on the relationship of the token to at least a portion of the remainder of the corpus. A set of sentiment scores may be obtained. The sentiment scores in the set of sentiment scores may describe a sentiment associated with a corresponding token that is described by a vector from the set of vectors. The set of vectors and the set of sentiment scores may be input into a pattern-recognizer pathway in a first neural network. A probability value of a potential future event may then be generated by the first neural network. The probability value may be based on the set of vectors and the set of sentiment scores.

Abstract: A machine learning method based on weakly supervised learning according to an embodiment of the present invention includes extracting feature maps about a dataset given a first type of information and not given a second type of information by using a convolutional neural network (CNN), updating the CNN by back-propagating a first error value calculated as a result of performing a task corresponding to the first type of information by using a first model, and updating the CNN by back-propagating a second error value calculated as a result of performing the task corresponding to the first type of information by using a second model different from the first model, wherein the second type of information is extracted when the task corresponding to the first type of information is performed using the second model.

Abstract: A method of updating a neural network may be provided. A method may include selecting a number of neurons for a layer for a neural network such that the number of neurons in the layer is less than at least one of a number of neurons in a first layer of the neural network and a number of neurons in a second, adjacent layer of the neural network. The method may further include and at least one of inserting the layer between the first layer and the second layer of the neural network and replacing one of the first layer and the second layer with the layer to reduce a number of connections in the neural network.

Abstract: Embodiments of the present invention provide an artificial neural network system for feature pattern extraction and output labeling. The system comprises a first spiking neural network and a second spiking neural network. The first spiking neural network is configured to autonomously learn complex, temporally overlapping features arising in an input pattern stream. Competitive learning is implemented as spike timing dependent plasticity with lateral inhibition in the first spiking neural network. The second spiking neural network is connected with the first spiking neural network through dynamic synapses, and is trained to interpret and label the output data of the first spiking neural network. Additionally, the labeled output of the second spiking neural network is transmitted to a computing device, such as a central processing unit for post processing.

Abstract: A device may receive issue resolution information, associated with a cognitive model, including an item of issue resolution information that describes an issue and a resolution to the issue. The device may assign the item of issue resolution information to a domain hierarchy, where the assigning is associated with a first user. The device may generate a question and an answer corresponding to the item of issue resolution information, where the generating of the question and the answer is associated with a second user. The device may approve the question and the answer, where the approving is associated with a third user. The device may generate a question/answer (QA) pair for the question and the answer. The device may create a data corpus including the QA pair, and provide the data corpus to cause the cognitive model to be trained based on the data corpus.

Abstract: In configuring a processing system with an application made up of machine learning components, where the application has been trained on a set of training data, the application is executed on the processing system using another set of training data. Outputs of the application produced from the other set of training data identified that concur with ground truth data are identified. The components are adapted to produce outputs of the application that concur with the ground truth data using the identified outputs of the application.

Abstract: A method for hardware-implemented training of a feedforward artificial neural network is provided. The method comprises: generating a first output signal by processing an input signal with the network, wherein a cost quantity assumes a first cost value; measuring the first cost value; defining a group of at least one synaptic weight of the network for variation; varying each weight of the group by a predefined weight difference; after the variation, generating a second output signal from the input signal to measure a second cost value; comparing the first and second cost values; and determining, based on the comparison, a desired weight change for each weight of the group such that the cost function does not increase if the respective desired weight changes are added to the weights of the group. The desired weight change is based on the weight difference times ?1, 0, or +1.

Abstract: Some embodiments perform, in a multi-layer neural network in a computing device, a convolution operation on input feature maps with multiple convolutional filters. The convolutional filters have multiple filter precisions. In other embodiments, electronic design automation (EDA) systems, methods, and computer-readable media are presented for adding such a multi-layer neural network into an integrated circuit (IC) design.

Abstract: In one aspect, a request for web content is received from a user device communicatively coupled to the processing device via the network. In response to receiving the request, user information associated with the user is determined. Predicted responses of the user to each variation of a plurality of variations of the web content are determined using prediction models and the user information. The prediction models include one or more decision trees generated using a splitting criterion requiring a minimum number of positive responses to a variation and a minimum number of negative responses to the variation as a condition of considering the possible split. The variation determined to have a threshold likelihood of yielding a predicted positive response of the predicted responses is selected based on the user information. The variation is transmitted to the user device via the network.

Abstract: Embodiments provide a computer implemented method, in a data processing system including a processor and a memory including instructions which are executed by the processor to cause the processor to train an enhanced chatflow system, the method including: ingesting a corpus of information including at least one user input node corresponding to a user question and at least one variation for each user input node; for each user input node: designating the node as a class; storing the node in a dialog node repository; designating each of the at least one variations as training examples for the designated class; converting the classes and the training examples into feature vector representations; training one or more training classifiers using the one or more feature vector representations of the classes; and training classification objectives using the one or more feature vector representations of the training examples.

Abstract: Statistical modeling techniques based neural network models for generating intelligence reports is provided. The system obtains test dataset and training dataset, each of which include at least one of images and elements. Statistical modeling techniques are identified and selected based on the test dataset for normalizing the test dataset to obtain normalized dataset. The system further associates, using one or more clustering techniques a unique cluster head to at least one of (i) normalized elements set and (ii) normalized images set in the normalized dataset to obtain a labeled dataset. The labeled dataset is further analysed by integrated trained modeling techniques into neural network model(s) and intelligence reports are generated.

Abstract: A method includes receiving, by a processor, bias data categories. A data input from a user for classification in data categories is received. A classification machine learning model is utilized to classify the data input in at least one data category and determine a first confidence probability in a classification outcome. A bias filter machine learning model is utilized to determine a second confidence probability that the classification outcome of classifying the data input into the at least one data category is based on at least one bias characteristic associated with at least one bias data category. A gate machine learning model is utilized to determine when to output the classification outcome of classifying the data input into the at least one data category to a computing device of a user based at least in part on the first confidence probability, the second confidence probability, and a predefined bias threshold.

Abstract: According to one embodiment, a learning apparatus includes a first rule generator, a feature value calculator, a related word extractor, a second rule generator, and a learning unit. The first rule generator generates a first rule to label the event candidate, the first rule including a keyword of the event candidate. The feature value calculator calculates feature values of other words included in the text other than the event candidate. The related word extractor extracts a related word relating to the keyword from the other words using the feature values. The second rule generator generates a second rule to label the event candidate, the second rule being different from the first rule and including the related word. The learning unit generates learning data associating the keyword, the related word, and labeled event candidate with each other.

Abstract: Disclosed herein is a denoising device including a deriving part configured to, when corrupted noise data corrupted due to noises is received from source data, derive an estimated loss which is estimated when each symbol within noise data is reconstructed to the source data based on a predefined noise occurrence probability, a processor to process training of a defined learning model by including parameters related with the reconstruction of the source data from the noise data based on context composed of a sequence of neighbored symbols based on each symbol within the noise data and pseudo-training data using the estimated loss corresponding to the context, and an output part to output reconstructed data in which each symbol within the noise data is reconstructed to a symbol of the source data through a denoiser formed based on a result of the training processing.