Abstract: A machine learning apparatus that learns laser machining condition data of a laser machining system includes: a state amount observation unit that observes a state amount of the laser machining system; an operation result acquisition unit that acquires a machined result of the laser machining system; a learning unit that receives an output from the state amount observation unit and an output from the operation result acquisition unit, and learns the laser machining condition data in association with the state amount and the machined result of the laser machining system; and a decision-making unit that outputs laser machining condition data by referring to the laser machining condition data learned by the learning unit.

Abstract: A system for using hash keys to preserve privacy across multiple tasks is disclosed. The system may provide training batch(es) of input observations each having a customer request and stored task to an encoder, and assign a hash key(s) to each of the stored tasks. The system may provide a new batch of input observations with a new customer request and new task to the encoder. The encoder may generate a new hash key assigned to the new customer request and determine whether any existing hash key corresponds with the new hash key. If so, the system may associate the new batch of input observations with the corresponding hash key and update the corresponding hash key such that it is also configured to provide access to the new batch of input observations. If not, the system may generate a new stored task and assign the new hash key to it.

Abstract: The present invention is directed to a system and methods for integrating flexible event detection into a near real-time streaming environment. A streaming platform streams raw data from multiple sources, adds contextual information to the raw data, and makes inferences from the enriched information. A rule execution operator, being integrated within the streaming platform, executes rules against the enriched information to detect events, using a rule repository that stores a plurality of flexible, dynamic, and customizable rules. When an event is detected by the rule execution operator, the streaming platform may use a middleware component to instantiate actions that are responsive to a detected event. Actions may be directly performed, and/or the actions and instructions for implementing or performing those actions are communicated to external devices.

Abstract: A method and system for improving analysis of social media and other usage data to determine user sentiments are disclosed. Social media posts are identified as relevant to determining user sentiments regarding a service provider. Posts are analyzed by machine learning algorithms to determine user general sentiments and specific sentiments. User interaction metrics indicating user interaction with service provider web site or application may also be analyzed. Sentiment and interaction determinations may be used with other data to predict likelihood of user attrition for services of the service provider. Sentiment determinations associated with social media posts may further be used to determine priority levels for the posts, including response urgency levels. Determined priority levels may then be used to implement appropriate actions in a timely manner based upon the post urgency.

Abstract: A learning use data set showing relationships among an engine speed, an engine load rate, an air-fuel ratio of the engine, an ignition timing of the engine, an HC or CO concentration of exhaust gas flowing into an exhaust purification catalyst and a temperature of the exhaust purification catalyst is acquired. The acquired engine speed, engine load rate, air-fuel ratio of the engine, ignition timing of the engine, and HC or CO concentration of the exhaust gas flowing into the exhaust purification catalyst are used as input parameters of a neural network and the acquired temperature of the exhaust purification catalyst is used as training data to learn a weight of the neural network. The learned neural network is used to estimate the temperature of the exhaust purification catalyst.

Abstract: Techniques described herein are directed to improved artificial neural network machine learning techniques that may be employed with a recommendation system to provide predictions with improved accuracy. In some embodiments, item consumption events may be identified for a plurality of users. From these item consumption events, a set of inputs and a set of outputs may be generated according to a data split. In some embodiments, the set of outputs (and potentially the set of inputs) may include item consumption events that are weighted according to a time-decay function. Once a set of inputs and a set of outputs are identified, they may be used to train a prediction model using an artificial neural network. The prediction model may then be used to identify predictions for a specific user based on user-specific item consumption event data.

Abstract: An agricultural data collection framework is provided in a system and method for tracking and managing livestock, and for analyzing animal conditions such as health, growth, nutrition, and behavior. The framework uses ultra-high frequency interrogation of RFID tags to collect individual animal data across multiple geographical locations, and incorporates artificial intelligence techniques to develop machine learning base models for statistical process controls around each animal for evaluating the animal condition. The framework provides a determination of normality at an individual animal basis or for a specific location, and generates alerts, predictions, and a targeted processing or application schedule for prioritizing and delivering resources when intervention is needed.

Abstract: Method, system, and apparatus for automatic stance classification. Propositions can be collected that are relevant to a query. A classifier can classify the stance of each proposition based on whether the proposition supports the query, opposes the query, or is neutral with respect to the query in order to thereafter provide substantive data for decision making based on and extracted from the query. The stance can be classified based on, for example, an SVM-SC (SVM Based Stance Classification) approach and/or an NN-SC (Neural Network Stance Classification Approach).

Abstract: A system for predicting that a user session will be fraudulent. The system can analyze an incomplete session and determine the likelihood that the session is fraudulent or not by generating completed sessions based on the incomplete session.

Abstract: A method for learning an adaption network corresponding to an obfuscation network used for concealing original data is provided. The method includes steps of: (a) on condition that a 1-st learning device has performed or is performing processes of (i) instructing the obfuscation network to obfuscate the training data to generate obfuscated training data, (ii) inputting the obfuscated training data into a learning network to generate 1-st characteristic information for training and inputting the training data into the learning network to generate 2-nd characteristic information for training, and (iii) learning the obfuscation network, a 2-nd learning device performing one of inputting the training data into the adaption network to generate 1-st feature adapted data and inputting test data into the adaption network to generate 2-nd feature adapted data and one of (i) acquiring a 1-st adaption ground truth and learning the adaption network and (ii) learning the adaption network.

Abstract: A method for learning a user learning network to recognize obfuscated data created by concealing original data is provided. The method includes steps of: a 2-nd learning device, (a) on condition that a 1-st learning device has performed (i) instructing the obfuscation network to generate obfuscated training data, (ii) inputting (ii-1) the obfuscated training data into, to generate 1-st characteristic information for training, and (ii-2) the training data, to generate 2-nd characteristic information for training, into a learning network for training and (iii) learning the obfuscation network, and acquiring (i) the obfuscated training data and a training data GT, or (ii) obfuscated test data and a test data GT; (b) inputting (i) the obfuscated training data, to generate 3-rd characteristic information for training, or (ii) the obfuscated test data, to generate 4-th characteristic information for training, into the user learning network; and (c) learning the user learning network.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting actions to be performed by an agent interacting with an environment. In one aspect, a system comprises a grid cell neural network and an action selection neural network. The grid cell network is configured to: receive an input comprising data characterizing a velocity of the agent; process the input to generate a grid cell representation; and process the grid cell representation to generate an estimate of a position of the agent in the environment; the action selection neural network is configured to: receive an input comprising a grid cell representation and an observation characterizing a state of the environment; and process the input to generate an action selection network output.

Abstract: To provide a learning model construction device, abnormality detection device, abnormality detection system and server for performing abnormality detection using sound information of the surroundings of a production apparatus. A learning model construction device includes a voice acquisition unit that acquires voice data including the voice of an operator located in the vicinity of a production apparatus, via a mic; a label acquisition unit that acquires an abnormality degree related to a production line including the production apparatus as a label; and a learning unit that constructs a learning model for the abnormality degree, by performing supervised learning with a group of voice data and label as training data.

Abstract: A computing system trains a clustering model. A responsibility parameter vector is initialized for each observation vector that includes a probability value of a cluster membership in each cluster. (A) Beta distribution parameter values are computed for each cluster. (B) Parameter values are computed for a normal-Wishart distribution for each cluster. (C) Each responsibility parameter vector defined for each observation vector is updated using the computed beta distribution parameter values, the computed parameter values for the normal-Wishart distribution, and a respective observation vector of the plurality of observation vectors. (D) A convergence parameter value is computed. (E) (A) to (D) are repeated until the computed convergence parameter value indicates the responsibility parameter vector defined for each observation vector is converged. A cluster membership is determined for each observation vector using a respective, updated responsibility parameter vector.

Abstract: A set of data is received at a server from a corresponding set of mobile devices. At the server, a first data in the set of data is analyzed, the first data being received from a first device in the set of mobile devices, the analyzing detecting an event of interest in the first data, the event occurring in a first geographical area. Using the event, a future event is predicted in a second geographical area. A target area is computed, where the target area includes a location where the future event is likely to occur. A notification about the future event is sent to a second set of devices located in the target area.

Abstract: Disclosed herein are systems, devices, and methods related to assets and predictive models and corresponding workflows that are related to the operation of assets. In particular, examples involve assets configured to receive and locally execute predictive models, locally individualize predictive models, and/or locally execute workflows or portions thereof.

Abstract: Systems, methods, and apparatus, including computer programs encoded on a computer storage medium, for selecting an actions from a set of actions to be performed by an agent interacting with an environment. In one aspect, the system includes a dueling deep neural network. The dueling deep neural network includes a value subnetwork, an advantage subnetwork, and a combining layer. The value subnetwork processes a representation of an observation to generate a value estimate. The advantage subnetwork processes the representation of the observation to generate an advantage estimate for each action in the set of actions. The combining layer combines the value estimate and the respective advantage estimate for each action to generate a respective Q value for the action. The system selects an action to be performed by the agent in response to the observation using the respective Q values for the actions in the set of actions.

Abstract: A data set whose records include respective pairs of entity descriptors with at least some text and a representation of a relationship such as similarity between the entities of the pair is obtained. Using the data set, a neural network model is trained to generate relationship indicators for pairs of entity descriptors. In an extensible token model of the neural network model, a text token of a first attribute of a particular entity descriptor is represented by a plurality of features including a first feature which was added to the token model as a result of a programmatic request. A particular relationship indicator corresponding to a source entity descriptor and a target entity descriptor is generated using the trained neural network model.

Abstract: A data evaluation method includes a processor receiving data sets N, each of which has one or more parameters, applying the data sets N to a machine learning model and generating observations on the data sets N; and executing a network sensitivity analysis (NSA) that includes generating a N NSA curve for each of k distinct parameters in the N data sets including computing an observation ON with the data sets, and generating a N?j NSA curve for each of the of the k distinct parameters. Generating a N?J NSA curve includes removing the jth data set from the N data sets and computing an observation ONj. Executing the NSA further includes determining a contribution of a jth data set based on the k N NSA curves and the k N?j NSA curves; and computing a relative strength Sj of each of the N data sets.

Abstract: A method for determining specific conditions occurring on industrial equipment based upon received signal data from sensors attached to the industrial equipment is provided. Using a server computer system, signal data is received and aggregated into feature vectors. Feature vectors represent a set of signal data over a particular range of time. The feature vectors are clustered into subsets of feature vectors based upon attributes the feature vectors. One or more sample episodes are received, where a sample episode includes sample feature vectors and specific classification labels assigned to the sample feature vectors. A signal data model is created that includes the associated feature vectors, clusters, and assigned classification labels.