Abstract: A neural network system for generating a quality assurance alert is provided. A computing device analyzes a quality assurance profile. A computing device arranges data in neurons of, at least, a first layer of a neural network. A computing device generates a threshold level of prediction of quality assurance based, at least in part, on output data from a neural network. A computing device applies output data from a neural network to a regression profile to determine a probability that a quality assurance issue will occur. A computing device generates a message that includes a quality assurance evaluation based, at least, on the determined probability that the quality assurance issue will occur.

Abstract: A computer-implemented method, computer program product, and system are provided for deep reinforcement learning to control a subject device. The method includes training, by a processor, a neural network to receive state information of a target of the subject device as an input and provide action information for the target as an output. The method further includes inputting, by the processor, current state information of the target into the neural network to obtain current action information for the target. The method also includes correcting, by the processor, the current action information minimally to obtain corrected action information that meets a set of constraints. The method additionally includes performing an action by the subject device based on the corrected action information for the target to obtain a reward from the target.

Abstract: Disclosed herein are system, method, and computer program product embodiments for an improved spiking neural network (SNN) configured to learn and perform unsupervised extraction of features from an input stream. An embodiment operates by receiving a set of spike bits corresponding to a set synapses associated with a spiking neuron circuit. The embodiment applies a first logical AND function to a first spike bit in the set of spike bits and a first synaptic weight of a first synapse in the set of synapses. The embodiment increments a membrane potential value associated with the spiking neuron circuit based on the applying. The embodiment determines that the membrane potential value associated with the spiking neuron circuit reached a learning threshold value. The embodiment then performs a Spike Time Dependent Plasticity (STDP) learning function based on the determination that the membrane potential value of the spiking neuron circuit reached the learning threshold value.

Abstract: A method for training an adversarial generator from a data set and a classifier includes: (A) training a classical noise generator whose input includes an output of a quantum generator, the classical noise generator having a first set of parameters, the training comprising: sampling from the data set to produce a first sample and a first corresponding label for the first sample; producing an output of the classical noise generator based on the output of the quantum generator and the first sample; producing a noisy example based on the output of the classical noise generator and the first sample; providing the noisy example to the classifier to produce a second corresponding label for the first sample; updating the first set of parameters such that the first corresponding label of the first sample differs from the second corresponding label of the first sample.

Abstract: Storage devices each hold corresponding one of n weight coefficient groups obtained by dividing weight coefficients such that each group includes weight coefficients about at least two bits. Bit value calculation circuits each output a result (flag information) by determining whether to accept updating about each of the bits based on the weight coefficient group, a value of an updated bit, identification information, and thermal excitation energy and an updated value of an accepted bit whose uprate has been accepted. First selection circuits each select an accepted bit based on the flag information and output a state signal including the flag information, the updated value, and identification information associated with the accepted bit. A second selection circuit determines the updated bit based on the flag information in the state signal and supplies the value of the updated bit and the identification information to each of optimization apparatuses.

Abstract: Neural Networks such as Deep Neural Networks (DNNs) output calibrated probabilities that substantially represent frequencies of occurrences of events. A DNN propagates uncertainty information of a unit of the DNN from an input to an output of the DNN. The uncertain information measures a degree of consistency of the test data with training data used to train a DNN. The uncertainty information of all units of the DNN can be propagated. Based on the uncertainty information, the DNN outputs probability scores that reflect received input data that is substantially different from the training data.

Abstract: A computer-implemented method is provided for learning an action policy. The method includes obtaining, by a processor, environment dynamics including triplets of a state, an action, and a next state. The state in each of the triplets is an expert state. The method further includes training, by the processor using the environment dynamics as training data, a dynamics model which obtains a pair of the state and the action as an input and outputs, for each next state, state-transition probabilities. The method also includes learning, by the processor, the action policy using trajectories of expert states according to a supervised learning technique by back-propagating error gradients through the trained dynamics model.

Abstract: A system for classification to prognostic labels using expert inputs includes a classification device. The classification device is designed and configured to record at least a physiological input pertaining to a human subject, receive at least an expert submission pertaining to the human subject, the at least an expert submission including at least a diagnostic constraint, and transmit at least a diagnostic output to a client device. The system includes a machine-learning module operating on the classification device, the machine-learning module designed and configured to receive training data relating physiological input data to diagnostic data and generate at least a diagnostic output using machine learning as a function of the training data, the at least an expert submission and the at least a physiological input.

Abstract: A method to grow an artificial neural network is disclosed. A seed neural network is trained on all classes in a dataset. All classes in the dataset are applied to the seed network, and average output values of the seed network are calculated. Class members that are nearest to and furthest from the average output values are selected, the class members are applied to the seed network, and a standard deviation is calculated. Perceptrons are added to the seed network, and inputs of the added perceptrons are connected to the seed layer based on the calculated standard deviation. A classifier is then added to the outputs of the added perceptrons, and the seed network and the added perceptrons are trained using all members in the dataset.

Abstract: The present disclosure provides a method of recognising a first action. The method comprises determining a weighted combination of a first plurality of feature responses associated with a sequence of second actions over a period of time using a set of weights at a particular time instance in the period of time. The method then recognises the first action by processing, using a neural network, the weighted combination of the first plurality of feature responses and temporal position values of each of the first plurality of feature responses associated with the sequence of second actions.

Abstract: Embodiments disclosed herein describe systems, methods, and products that generate trained neural networks that are robust against adversarial attacks. During a training phase, an illustrative computer may iteratively optimize a loss function that may include a penalty for ill-conditioned weight matrices in addition to a penalty for classification errors. Therefore, after the training phase, the trained neural network may include one or more well-conditioned weight matrices. The one or more well-conditioned weight matrices may minimize the effect of perturbations within an adversarial input thereby increasing the accuracy of classification of the adversarial input. By contrast, conventional training approaches may merely reduce the classification errors using backpropagation, and, as a result, any perturbation in an input is prone to generate a large effect on the output.

Abstract: The disclosed computer-implemented method may include (1) identifying an artificial neural network comprising a set of nodes interconnected via a set of connections, and (2) training the artificial neural network by, for each connection in the set of connections, determining a quantized weight value associated with the connection. Determining the quantized weight value associated with the connection may include (1) associating a loss function with the connection, the loss function including a periodic regularization function that describes a relationship between an input value and a weight value of the connection, (2) determining a minimum of the associated loss function with respect to the weight value in accordance with the periodic regularization function, and (3) generating the quantized weight value associated with the connection based on the determined minimum of the loss function. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: In some aspects, a computing system can generate and optimize a neural network for risk assessment. The neural network can be trained to enforce a monotonic relationship between each of the input predictor variables and an output risk indicator. The training of the neural network can involve solving an optimization problem under a monotonic constraint. This constrained optimization problem can be converted to an unconstrained problem by introducing a Lagrangian expression and by introducing a term approximating the monotonic constraint. Additional regularization terms can also be introduced into the optimization problem. The optimized neural network can be used both for accurately determining risk indicators for target entities using predictor variables and determining explanation codes for the predictor variables. Further, the risk indicators can be utilized to control the access by a target entity to an interactive computing environment for accessing services provided by one or more institutions.

Abstract: An output value of a first neural network for input data is obtained in correspondence with each category. An output value of a second neural network generated by changing a designated unit in the first neural network is obtained for the input data in correspondence with each category. For each category, change information representing a change in the output value is obtained. Information representing contribution of the designated unit are output to a display device based on the change information.

Abstract: A micro-processor circuit and a method of performing neural network operation are provided. The micro-processor circuit is suitable for performing neural network operation. The micro-processor circuit includes a parameter generation module, a compute module and a compare logic. The parameter generation module receives in parallel a plurality of input parameters and a plurality of weight parameters of the neural network operation. The parameter generation module generates in parallel a plurality of sub-output parameters according to the input parameters and the weight parameters. The compute module receives in parallel the sub-output parameters. The compute module sums the sub-output parameters to generate a summed parameter. The compare logic receives the summed parameter. The compare logic performs a comparison operation based on the summed parameter to generate a plurality of output parameters of the neural network operation.

Abstract: Method includes recording biomarker information being indicative of at least one biological state of user remaining in lighting control environment over time frame, biomarker information being generated by at least one physiological sensor remaining with user in lighting control environment over time frame; recording light control settings for at least one light remaining in lighting control environment with user and with physiological sensor generating biomarker information over time frame; and training a neural network to determine correlations between biological state of user remaining in lighting control environment over time frame and lighting effects caused by at least one light remaining in lighting control environment with user over time frame, based on recordings of biomarker information and recordings of light control settings, and utilizing the correlations for controlling the at least one light. Computer readable medium for executing method.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating output examples using neural networks. One of the methods includes receiving a request to generate an output example of a particular type, accessing dependency data, and generating the output example by, at each of a plurality of generation time steps: identifying one or more current blocks for the generation time step, wherein each current block is a block for which the values of the bits in all of the other blocks identified in the dependency for the block have already been generated; and generating the values of the bits in the current blocks for the generation time step conditioned on, for each current block, the already generated values of the bits in the other blocks identified in the dependency for the current block.

Abstract: Sensor data captured by one or more sensors may be received at an analysis system. A neural network may be used to detect an object in the sensor data. A plurality of polygons surrounding the object may be generated in one or more subsets of the sensor data. A prediction of a future position of the object may be generated based at least in part on the polygons. One or more commands may be provided to a control system based on the prediction of the future position.

Abstract: A method for detecting an anomalous image among a dataset of images using an Adversarial Autoencoder includes training an Adversarial Autoencoder in a first training with a training dataset of images, with the Adversarial Autoencoder being optimized such that a distribution of latent representations of images of the training dataset of images approaches a predetermined prior distribution and that a reconstruction error of reconstructed images of the training dataset of images is minimized. Subsequently, anomalies are detected in the latent representation and the Adversarial Autoencoder is trained in a second training with the training dataset of images, but taking into account the detected anomalies. The anomalous image among the first dataset of images is detected by the trained Adversarial Autoencoder dependent on at least one of the reconstruction error of the image and a probability density under the predetermined prior distribution.

Abstract: Hardware optimization of neural networks is provided. In various embodiments, an output-induced receptive field of each of a plurality of layers of a neural network is determined. From each of the plurality of layers any portions of their respective input that falls outside their respective output-induced receptive field are trimmed. For each of the plurality of layers, a plurality of mappings of the layer to physical neurosynaptic cores are determined. A mapping is determined having a minimum total number of cores required for the neural network based on the plurality of mappings.