Abstract: A method and system are provided. The method includes receiving by a computer having a processor and a memory, sequence data that includes labeled data and unlabeled data. The method further includes generating, by the computer having the processor and the memory, a recurrent neural network model of the sequence data, the recurrent neural network model having a recurrent layer and an aggregate layer. The recurrent neural network model feeds sequences generated from the recurrent layer into the aggregate layer for aggregation, stores temporal dependencies in the sequence data, and generates labels for at least some of the unlabeled data.

Abstract: A system and method for a neural network that is trained to recognize patterns in the exitance convergence behaviour of a radiosity equation being solved for a set of finite element environments, and subsequently employed to monitor and predict the exitance convergence behaviour of novel finite element environments. The neural network is trained with feature vectors representing partial snapshots of exitance vectors at various iterations in a radiosity calculation. The feature vectors are related to numbers of iterations that can be skipped by making approximate calculations instead of performing the iterations. In use, when a radiosity equation is being solved, the neural network identifies feature vectors generated during the calculations that signify that a certain number of iterations can be skipped by making an approximate calculation.

Abstract: A sentiment analysis computing system includes a storage medium and a processing system. Sentiment input is received from audience members viewing a streamed/webcasted event. The received input is stored to the storage medium. A time slice of the webcasted event is determined and sentiment inputs that are within that time slice are obtained. A sentiment value is calculated for the determined time slice based on aggregated sentiment values. The calculated sentiment value for the time slice is then output by the sentiment analysis computing system.

Abstract: The present disclosure is directed to a high-capacity training and prediction machine learning platform that can support high-capacity parameter models (e.g., with 10 billion parameters). The platform generates a model for a metric of interest based on a known training set. The model includes parameters indicating importances of different features of the model, taken both singly and in pairs. The model may be applied to predict a value for the metric for given sets of objects, such as for a pair consisting of a user object and a content item object.

Abstract: Controllable resistance elements and methods of setting the same include a junction field effect transistor configured to provide a resistance on a signal line. A first pass transistor is configured to apply a charge increment or decrement to the junction field effect transistor responsive to a control pulse, such that the resistance on the signal line changes.

Abstract: Methods of training a neural network include applying an input signal to an array of weights to generate weighted output signals based on resistances of respective weights in the array of weights. A difference between the weighted output signals and a predetermined expected output is determined. Weights in the array of weights are set by applying a pulse to a controllable resistance element in each weight. The pulse increments or decrements a charge on a junction field effect transistor in the respective controllable resistance element.

Abstract: Neural networks include neuron layers arranged in order from an input neuron layer to an output neuron layer, with at least one hidden layer between them. Weight arrays between respective pairs of neuron layers each include controllable resistance elements and AND gates configured to control addressing of the plurality of controllable resistance elements. Each controllable resistance element includes a junction field effect transistor configured to provide a resistance on a signal line and a first pass transistor configured to apply a charge increment or decrement to the junction field effect transistor responsive to a control pulse, such that the resistance on the signal line changes. The control pulse is only passed to a controllable resistance element when a respective AND gate is triggered. A training module is configured to train the neural network by adjusting resistances of the plurality of controllable resistance elements in each of the weight arrays.

Abstract: Techniques and constructs can reduce the time required to determine solutions to optimization problems such as training of neural networks. Modifications to a computational model can be determined by a plurality of nodes operating in parallel. Quantized modification values can be transmitted between the nodes to reduce the volume of data to be transferred. The quantized values can be as small as one bit each. Quantization-error values can be stored and used in quantizing subsequent modifications. The nodes can operate in parallel and overlap computation and data transfer to further reduce the time required to determine solutions. The quantized values can be partitioned and each node can aggregate values for a corresponding partition.

Abstract: A method reads data from a synapse which includes a transistor and a variable resistor. The transistor has a gate electrode, a first electrode and a second electrode. The variable resistor has a first electrode connected to the second electrode of the transistor. The method includes applying a read voltage to the gate electrode of the transistor, applying a pre-synaptic voltage to the first electrode of the transistor, and applying a post-synaptic voltage to a second electrode of the variable resistor. The read voltage is lower than the threshold voltage of the transistor.

Abstract: Methods and apparatus are provided for estimating anomalous sensors. The apparatus includes a target data acquiring section to acquire a plurality of sets of target data serving as an examination target, each set of target data being output by a plurality of sensors. The apparatus further includes a calculating section to calculate, for each of a plurality of sensor groups such that each sensor group includes at least two sensors among the plurality of sensors, a degree of difference of a target data distribution of the plurality of sets of target data relative to a reference data distribution of output from the sensor group. The apparatus additionally includes an estimating section to estimate one or more sensors among the plurality of sensors to be a source of outlierness, based on a calculation result of the calculating section.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for implementing a convolutional gated recurrent neural network (CGRN). In one of the systems, the CGRN is configured to maintain a state that is a tensor having dimensions x by y by m, wherein x, y, and m are each greater than one, and for each of a plurality of time steps, update a currently maintained state by processing the currently maintained state through a plurality of convolutional gates.

Abstract: Techniques to mimic a target food item using artificial intelligence are disclosed. A formula generator is trained using combinations of ingredients. A training set may include, for each combination of ingredients, proportions, and features of the ingredients in a respective combination of ingredients. Given a target food item, the formula generator determines a predicted formula that matches the given target food item. The predicted formula includes a set ingredients and a respective proportion of each ingredient in the set of ingredient.

Abstract: From a sequence of answers, a last remaining answer is selected. A set of answers in the sequence of answers are responsive to a set of questions resolved during an analysis of a reported problem in a data processing environment. An answer pair is formed using the last remaining answer and another answer which immediately precedes the last remaining answer in the sequence of answers. A probability is determined of the last remaining answer being caused by the other answer in the answer pair. When the probability is below a threshold value, a review workflow is triggered corresponding to a portion of the analysis. The portion includes a question corresponding to an answer in the answer pair.

Abstract: Embodiments include methods and systems for using, creating and maintaining goal-oriented, dialog systems (i.e., transactional bots) that provide interfaces to application functionality. The methods and systems of the embodiments provide a bot that may learn in supervised learning and reinforcement learning from conversational examples provided by domain experts and from interaction with users. Conversational bots may be created to interact using both text and/or application programming interface (API) calls. A developer may configure a bot that interfaces with an application back-end where behavior of the bot may be controlled by use of masking actions. A specification for the bot may be flexibly designed to specify how developer code may be organized, for example, as masking operations on the possible actions the bot may execute. Additionally, the methods and systems may automatically infer the best state representation during a dialog so a state variable need not be predefined.

Abstract: A system and method provides efficient parallel training of a neural network model on multiple graphics processing units. A training module reduces the time and communication overhead of gradient accumulation and parameter updating of the network model in a neural network by overlapping processes in an advantageous way. In a described embodiment, a training module overlaps backpropagation, gradient transfer and accumulation in a Synchronous Stochastic Gradient Decent algorithm on a convolution neural network. The training module collects gradients of multiple layers during backpropagation of training from a plurality of graphics processing units (GPUs), accumulates the gradients on at least one processor and then delivers the gradients of the layers to the plurality of GPUs during the backpropagation of the training. The whole model parameters can then be updated on the GPUs after receipt of the gradient of the last layer.

Abstract: A machine learning device of a finish-machining amount prediction apparatus observes, as state variables expressing a current state of an environment, finish-machining amount data indicating finish-machining amounts of the respective parts of a component and accuracy data indicating the accuracy of the respective parts of a machine, to which the component is attached. Then, the machine learning device acquires determination data indicating propriety determination results of the accuracy of the respective parts of the machine, to which the component after being subjected to finish machining is attached. After that, the machine learning device learns the finish-machining amounts of the respective parts of the component in association with the accuracy data by using the state variables and the determination data.

Abstract: A method is disclosed including presenting a concept to a user via one or more presentation devices and monitoring the user's response to the presentation of the concept by a sensing device. The sensing device may generate sensor data based on the monitored user's response. The method further includes determining based on the sensor data generated by the sensor that the user requires clarification of the presented concept. In response to determining that the user requires clarification of the presented concept, the method further includes identifying an analogy that is configured to clarify the presented concept and presenting the identified analogy to the user via one or more of the presentation devices.

Abstract: Techniques to mimic a target food item using artificial intelligence are disclosed. A formula generator learns from open source and proprietary databases of ingredients and recipes. The formula generator is trained using features of the ingredients and using recipes. Given a target food item, the formula generator determines a formula that matches the given target food item and a score for the formula. The formula generator may generate numerous formulas that match the given target food item and may select an optimal formula from the generated formulas based on score.

Abstract: A processing unit can determine a first feature value corresponding to a session by operating a first network computational model (NCM) based part on information of the session. The processing unit can determine respective second feature values corresponding to individual actions of a plurality of actions by operating a second NCM. The second NCM can use a common set of parameters in determining the second feature values. The processing unit can determine respective expectation values of some of the actions of the plurality of actions based on the first feature value and the respective second feature values. The processing unit can select a first action of the plurality of actions based on at least one of the expectation values. In some examples, the processing unit can operate an NCM to determine expectation values based on information of a session and information of respective actions.

Abstract: In one embodiment, a device trains a machine learning-based malware classifier using a first randomly selected subset of samples from a training dataset. The classifier comprises a random decision forest. The device identifies, using at least a portion of the training dataset as input to the malware classifier, a set of misclassified samples from the training dataset that the malware classifier misclassifies. The device retrains the malware classifier using a second randomly selected subset of samples from the training dataset and the identified set of misclassified samples. The device adjusts prediction labels of individual leaves of the random decision forest of the retrained malware classifier based in part on decision changes in the forest that result from assessing the entire training dataset with the classifier. The device sends the malware classifier with the adjusted prediction labels for deployment into a network.