Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating sequences of predicted observations, for example images. In one aspect, a system comprises a controller recurrent neural network, and a decoder neural network to process a set of latent variables to generate an observation. An external memory and a memory interface subsystem is configured to, for each of a plurality of time steps, receive an updated hidden state from the controller, generate a memory context vector by reading data from the external memory using the updated hidden state, determine a set of latent variables from the memory context vector, generate a predicted observation by providing the set of latent variables to the decoder neural network, write data to the external memory using the latent variables, the updated hidden state, or both, and generate a controller input for a subsequent time step from the latent variables.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for determining neural network architectures. One of the methods includes obtaining training data for a dense image prediction task; and determining an architecture for a neural network configured to perform the dense image prediction task, comprising: searching a space of candidate architectures to identify one or more best performing architectures using the training data, wherein each candidate architecture in the space of candidate architectures comprises (i) the same first neural network backbone that is configured to receive an input image and to process the input image to generate a plurality of feature maps and (ii) a different dense prediction cell configured to process the plurality of feature maps and to generate an output for the dense image prediction task; and determining the architecture for the neural network based on the best performing candidate architectures.

Abstract: Methods and systems for measuring and predicting content dissemination in social networks includes computing a “virality score” for popularity of social media content, a “pattern” of diffusion of the content, and a “hype” parameter of such content without requiring a “friendship graph” or “information diffusion” structured data. It also includes an iterative, predictive model, which predicts future performance (or future rate of dissemination) of the content while mitigating a class imbalance problem inherent in predicting viral posts, and which provides updates to the model based on actual performance results.

Abstract: A method and system for predicting system states is provided. The method includes receiving a first reference model associated with a first operational attribute of a system from a first integrated circuit internally comprising a first processing circuit and a first sensor measuring a first parameter of the system. Additionally, a second reference model associated with a second operational attribute of the system is received from a second integrated circuit internally comprising a second processing circuit and a second sensor measuring a second parameter of the system. A combination reference model based on the first reference model and the second reference model is generated and a predicted future state and associated operational attributes for the system are determined based on the combination reference model.

Abstract: A computer-implemented method is disclosed that operates in conjunction with machine learning binary classification of an event. The method generates a rule relating to whether or not the event will occur. In one embodiment, the method includes obtaining foreground sequences corresponding to the event happening and background sequences corresponding to the event not happening. For each foreground sequence of a number of foreground sequences, the computer computes a plurality of difference values. The rule may be determined based on a largest difference value for one of the plurality of foreground sequences. A corresponding system is also disclosed.

Abstract: A learning method for generating integrated object detection information by integrating first object detection information and second object detection information is provided. And the method includes steps of: (a) a learning device instructing a concatenating network to generate one or more pair feature vectors; (b) the learning device instructing a determining network to apply FC operations to the pair feature vectors, to thereby generate (i) determination vectors and (ii) box regression vectors; (c) the learning device instructing a loss unit to generate an integrated loss by referring to the determination vectors, the box regression vectors and their corresponding GTs, and performing backpropagation processes by using the integrated loss, to thereby learn at least part of parameters included in the DNN.

Abstract: The present disclosure provides systems and methods for the management and/or evaluation of machine-learned models based on locally logged data. In one example, a user computing device can obtain a machine-learned model (e.g., from a server computing device) and can evaluate at least one performance metric for the machine-learned model. In particular, the at least one performance metric for the machine-learned model can be evaluated relative to data that is stored locally at the user computing device. The user computing device and/or the server computing device can determine whether to activate the machine-learned model on the user computing device based at least in part on the at least one performance metric. In another example, the user computing device can evaluate a plurality of machine-learned models against locally stored data. At least one of the models can be selected based on the evaluated performance metrics.

Abstract: Users may request assistance or information from a limited number of resources, such as submitting a user request by speaking or entering text. A user request from among the pending user requests may be selected using a selection model. A selection model may process features relating to each of the pending user requests and generate a probability distribution for the pending user requests. A user request may then be selected using the probability distribution, such as by making a random selection. The selection model may be updated over multiple time periods by computing reward scores for the selection decisions made by the selection model and using the reward scores to update the parameters of the selection model.

Abstract: A method for processing tensor data for pattern recognition and a computer device are provided. The method includes: constructing a decision function by the optimal projection tensor W which has been rank-one decomposed together with the offset scalar b, and inputting to-be-predicted tensor data which has been rank-one decomposed into the decision function for prediction.

Abstract: A neuromorphic memory circuit including a programmable resistive memory element, an axon LIF pulse generator to generate an axon LIF pulse, a back propagation pulse generator to generate a back propagation pulse, a postsynaptic capacitor configured to build up a forward propagation LIF charge over time, and a presynaptic capacitor configured to build up a back propagation LIF charge over time. A first transistor activates a first discharge path from the postsynaptic capacitor through the programmable resistive memory element when the axon LIF pulse generator generates the axon LIF pulse. A second transistor activates a second discharge path from the presynaptic capacitor through the programmable resistive memory element when the back propagation pulse generator generates the back propagation pulse.

Abstract: In one embodiment, a system retrieves a first feature vector for an image. The image is inputted into a first deep-learning model, which is a first-version model, and the first feature vector may be output from a processing layer of the first deep-learning model for the image. The first feature vector using a feature-vector conversion model to obtain a second feature vector for the image. The feature-vector conversion model is trained to convert first-version feature vectors to second-version feature vectors. The second feature vector is associated with a second deep-learning model, and the second deep-learning model is a second-version model. The second-version model is an updated version of the first-version model. A plurality of predictions for the image may be generated using the second feature vector and the second deep-learning model.

Abstract: The mirror deep neural networks (DNNs) as described herein recognize patterns in an input signal. Mirror DNNs regularize to a linear function and train very quickly. Mirror DNNs employ a neural network pattern recognizer that receives a set of features extracted from an input signal and inputs the set of features into a multi-layer neural network. The multi-layer neural network has an input layer that receives the set of features, a plurality of intermediate layers, and an output layer that generates a set of output values that are indicative of a recognized pattern exhibited in the input signal. A first and second non-linear equation pair are chosen and applied to intermediate layers of the neural network so as to make the output values that are indicative of a pattern exhibited in the input signal linear.

Abstract: A method for the coherent tracking of a time varying signal using evolutionary computing including global and local optimization techniques for the purpose of obtaining better performance under poor signal reception conditions, multipath errors, indoors, and for obtaining more accurate estimates of carrier phase, carrier frequency, and modulation phase at low signal levels without being subject to the traditional phase lock tracking loops (PLL) or delay lock tracking loops (DLL) limitations.

Abstract: Methods and systems for detecting new calls from an existing spam or robocaller and aggregating calls that originate from the same infrastructure using a telephony honeypot are disclosed. An example method may receive a telephone call to a telephony honeypot and store metadata and a call audio recording associated with the telephone call. A transcript may be created of the call audio recording. Audio features may be extracted from the call audio recording. The transcript may be compared to other previously-received transcripts in order to determine a similarity between the call and previously-received calls. Audio features and metadata may also be compared to determine whether the call is similar to other previously-received calls. If a call is similar, the call may be identified with the same spam campaign or robocaller as the similar, previously-received call.

Abstract: A decision support system is described that solves various technical problems associated with prior systems. The system may include a thin client application that downloads question sets from a server. The question sets may include a series of interlinked questions and answers as well as scores associated with the answers. Scores of selected answers may be tallied and a decision corresponding to the cumulative score of the selected answers may be provided to a user. As a result, the system may enable the ability to update (add, revise or delete) questions, answers and question lines without needing to recompile and redistribute the client application. In addition, the transmission of a complete question set reduces overall network traffic and technical problems arising from network disruptions among other technical benefits.

Abstract: A method, system and computer-usable medium for providing cognitive insights comprising receiving data from a plurality of data sources, the plurality of data sources comprising a blockchain data source, the blockchain data source providing blockchain data; processing the data from the plurality of data sources, the processing the data from the plurality of data sources performing data enriching to provide enriched data; generating the cognitive session graph, the cognitive session graph being associated with a session, the cognitive session graph comprising at least some enriched data; and, associating a cognitive blockchain with the cognitive session graph.

Abstract: A data architecture for use within a cognitive information processing system environment comprising: a plurality of data sources, the plurality of data sources comprising a public data source and a private data source, the public source comprising a public blockchain data source, the private data source comprising a private blockchain data source; a cognitive data management module, the cognitive data management module accessing information from the plurality of data sources, the information comprising public blockchain information and private blockchain information; and, providing the information to a cognitive inference and learning system.

Abstract: A method, system and computer-usable medium for providing cognitive insights comprising receiving streams of data from a plurality of data sources; receiving data from a plurality of data sources, the plurality of data sources comprising a blockchain data source, the blockchain data source providing blockchain data; processing the data from the plurality of data sources, the processing the data from the plurality of data sources performing data enriching to provide enriched data; generating the cognitive session graph, the cognitive session graph being associated with a session, the cognitive session graph comprising at least some enriched data; and, associating a cognitive blockchain with the cognitive session graph.

Abstract: A method for receiving a plurality of types of data within a cognitive information processing system environment comprising: receiving data from a plurality of data sources, the plurality of data sources comprising a public data source and a private data source the public source comprising a public blockchain data source, the private data source comprising a private blockchain data source; accessing information from the plurality of data sources via a cognitive data management module, the information comprising public blockchain information and private blockchain information; and, providing the information to an inference and learning system.

Abstract: Systems, apparatuses, and methods for the automated detection of incorrect data during a data entry process or detecting incorrect data that has been entered and stored previously. In one embodiment, the invention utilizes one or more of statistical analysis or a machine learning technique (either supervised or unsupervised) in order to identify potentially incorrect data.