Abstract: A computer-implemented system and method of analyzing spatial, temporal and contextual elements of data for predictive decision-making. The computer-implemented method includes receiving a first set of rules and receiving first data and second data including spatial, temporal and contextual elements. The computer-implemented method also includes identifying each rule of the first set for which the received first data and the received second data is a respective candidate. For the identified rules for which the received first data is a candidate, and for the identified rules for which the received second data is a candidate, the respective received first data and received second data is indexed by its temporal, spatial or contextual elements as a function of the identified rules. The computer-implemented method also includes identifying an event as satisfying an identified rule in memory using the indexed first and second data.

Abstract: The present description is directed towards systems and methods for directing a user request for content over a network to a given content server on the basis of one or more rules. Methods and systems implemented in accordance with the present description comprise receiving a request for content form a user, the request for content including a profile of the user identifying one or more characteristics associated with the user. One or more rules are retrieved for identifying a content server to which a request for content is to be delivered, the one or more rules including at least one of business rules, network rules, and user profile rules. The one or more retrieved rules are applied to the request for content to identify a content server to which the request for content is to be delivered and the request for content is delivered to the identified content server.

Abstract: A computer system and method for performing predictive analytics on telematics data regarding an entity. The computer system having a memory configured to store instructions and a processor disposed in communication with the memory. The processor upon execution of the instructions is configured to receive telematics data regarding an entity and analyze the received telematics data to identify a pattern of behavior. A behavioral conclusion and/or meaning is then determined for the entity based on analysis of the received telematics data.

Abstract: A method for using a structurally more complicated reference model to train a structurally simpler learner model includes: obtaining a trained reference model at least including N reference blocks and a learner model at least including N learner blocks respectively corresponding to the N reference blocks; training the learner model by conducting an iterative operation; determining whether the learner model is convergent; and in response to that the learner model is convergent, stopping the iterative operation to assign the learner model as a trained learner model. The iterative operation includes inputting a sample data set into the trained reference model and the learner model; for each of the N learner blocks: determining a distance between a learner vector of the learner block and a reference vector of the reference block, and updating parameters in the learner block based on the determined distance.

Abstract: A mobile application is provided that provides intelligent recommendations based on the knowledge of where the user has been, and what venues the user would like to visit. Further, such an application may be capable of determining where people in a user's social network have been and what venue locations these related users would like to visit. Also, in another implementation, the application may be capable of determining where people with similar taste have been, and where they would like to go. Some or all of this information may be used by a mobile application that provides recommendations to a user. For instance, in one implementation, a user having a mobile device such as a cell phone wishes to locate a venue based on one or more parameters, and some or all of this information may be used to order to rank recommendations with the interface.

Abstract: A machine learning apparatus determines an order in which numerical values in an input dataset are to be entered to a neural network for data classification, based on a reference pattern that includes an array of reference values to provide a criterion for ordering the numerical values. The machine learning apparatus then calculates an output value of the neural network whose input-layer neural units respectively receive the numerical values arranged in the determined order. The machine learning apparatus further calculates an input error at the input-layer neural units, based on a difference between the calculated output value and a correct classification result indicated by a training label. The machine learning apparatus updates the reference values in the reference pattern, based on the input error at the input-layer neural units.

Abstract: Automation security in a networked-based industrial controller environment is implemented. Various components, systems and methodologies are provided to facilitate varying levels of automation security in accordance with security analysis tools, security validation tools and/or security learning systems. The security analysis tool receives abstract factory models or descriptions for input and generates an output that can include security guidelines, components, topologies, procedures, rules, policies, and the like for deployment in an automation security network. The validation tools are operative in the automation security network, wherein the tools perform security checking and/or auditing functions, for example, to determine if security components are in place and/or in suitable working order.

Abstract: A medical diagnosis support apparatus which provides information for supporting medical diagnosis includes an inference unit which obtains an inference result based on a combination of already input information and each non-input information, an evaluation unit which evaluates each non-input information by using an inference result on the already input information which is obtained by the inference unit and an inference result on the each non-input information which is obtained by the inference unit, and a selection unit which selects non-input information to be presented from the non-input information based on the evaluation obtained by the evaluation unit.

Abstract: A system, method, and computer-readable medium are disclosed for performing a dynamic application optimization operation, comprising: instrumenting a plurality of system parameters of a client information handling system for monitoring; instructing a user to execute a particular application on the client information handling system; obtaining a plurality of samples of the plurality of system parameters; performing a machine learning operation using the plurality of samples of the plurality of system parameters, the machine learning operation training a machine learning model to generate a profile for the particular application and an operating mode of the particular application; applying the profile to the client information handling system to provide a new information handling system configuration, the new information handling system configuration optimizing the information handling system for the particular application.

Abstract: Systems, methods, and computer-readable storage media that may be used to generate a category Bayesian hierarchical model. One method includes receiving a brand data set for each of a plurality of brands within a category, each brand data set comprising content input for a particular brand of the plurality of brands for a plurality of media channels at a plurality of times and a response for the particular brand of the plurality of brands at the plurality of times. The method includes determining a plurality of informative priors by generating a category Bayesian hierarchical model based on the plurality of brand data sets and a plurality of weak priors. The method further includes generating a brand Bayesian hierarchical model that models response for the particular brand for each of the plurality of media channels based on the brand data set for the particular brand and the plurality of informative priors.

Abstract: Techniques are for rule-based continuous drift and consistency management for target systems. In one embodiment, a set of rules is stored in volatile or non-volatile store. The set of rules may include one or more drift rules and/or one or more consistency rules. A rule may be applied to one or more associated targets to detect drift or inconsistency. A drift rule identifies a set of one or more attributes and a source and may be applied by comparing a first configuration of the set of one or more attributes on an associated target with a second configuration of the set of one or more attributes on the source. A consistency rule may be applied to a composite target by comparing member targets that are grouped by target type. Notification data may be output if target drift or inconsistency is detected to alert a user.

Abstract: Certain aspects of the present disclosure provide systems and methods for configuring and training neural networks. The method includes models of individual neurons in a network that avoid certain biologically impossible or implausible features of conventional artificial neural networks. Exemplary networks may use patterns of local connections between excitatory and inhibitory neurons to provide desirable computational properties. A network configured in this manner is shown to solve a digit classification problem.

Abstract: A system for compressed data storage using a neural network. The system comprises a memory comprising a plurality of memory locations configured to store data; a query neural network configured to process a representation of an input data item to generate a query; an immutable key data store comprising key data for indexing the plurality of memory locations; an addressing system configured to process the key data and the query to generate a weighting associated with the plurality of memory locations; a memory read system configured to generate output memory data from the memory based upon the generated weighting associated with the plurality of memory locations and the data stored at the plurality of memory locations; and a memory write system configured to write received write data to the memory based upon the generated weighting associated with the plurality of memory locations.

Abstract: A programmable architecture specialized for convolutional neural networks (CNNs) processing such that different applications of CNNs may be supported by the presently disclosed method and apparatus by reprogramming the processing elements therein. The architecture may include an optimized architecture that provides a low-area or footprint and low-power solution desired for embedded applications while still providing the computational capabilities required for CNN applications that may be computationally intensive, requiring a huge number of convolution operations per second to process inputs such as video streams in real time.

Abstract: A method of identifying grains in polycrystalline materials, the method including (a) identifying local crystal structure of the polycrystalline material based on neighbor coordination or pattern recognition machine learning, the local crystal structure including grains and grain boundaries, (b) pre-processing the grains and the grain boundaries using image processing techniques, (c) conducting grain identification using unsupervised machine learning; and (d) refining a resolution of the grain boundaries.

Abstract: A neural network system includes an input layer, one or more hidden layers, and an output layer. The input layer receives a training set including a sequence of batches and provides to its following layer output activations associated with the sequence of batches respectively. A first hidden layer receives, from its preceding layer, a first input activation associated with a first batch, receive a first input gradient associated with a second batch preceding the first batch, and provide, to its following layer a first output activation associated with the first batch based on the first input activation and first input gradient. The first and second batches have a delay factor associated with at least two batches. The output layer receives, from its preceding layer, a second input activation, and provide, to its preceding layer, a first output gradient based on the second input activation and the first training set.

Abstract: A first signature log generated by a first processing system is received, the first signature log including a first listing of at least a recorded series of user inputs received by a computer program. The first listing of the recorded series of user inputs can be compared to at least one historical log indicating at least one historical path, each historical path including at least second listing of a series of historical user inputs that corresponds to a historical endpoint in the computer program. The at least one historical path to which the recorded series of user inputs at least partially corresponds and a deviation between the signature log and the historical log can be identified. Based on a predicted endpoint, a predicted path corresponding to the predicted endpoint can be selected, and the predicted path and the deviation between the signature log and the historical log can be output.

Abstract: One embodiment of the invention provides a system for mapping a neural network onto a neurosynaptic substrate. The system comprises a metadata analysis unit for analyzing metadata information associated with one or more portions of an adjacency matrix representation of the neural network, and a mapping unit for mapping the one or more portions of the matrix representation onto the neurosynaptic substrate based on the metadata information.

Abstract: Embodiments herein achieve a machine-learning system for managing shuffling of input training datasets. The machine-learning system includes a training dataset manager configured to shuffle an input dataset received from each of a plurality of electronic devices. Further, the training dataset manager is configured to split the input training datasets into a plurality of mini-batches. Each of the mini-batches, along with the target values, defines an error surface corresponding to an error function. A learning manager is configured to obtain a cross mini-batch discriminator based on the error function for each of the mini-batches. Further, the learning manager is configured to select a mini-batch associated with a least cross mini-batch discriminator from the plurality of mini-batches as optimal mini-batch.

Abstract: A networked interactive expert system is disclosed for selectively providing relevant expertise when a user requires such assistance. System provides a set of networked facilities for determining an expertise needed by a customer based on a location. Such location can be either physical or logical. A physical “location” corresponds, for example, to a department within a retail outlet. Examples of logical “locations” are ones corresponding to a web-page, a product identification code of interest, a customer identification code, and/or explicit knowledge category selected by the customer. An expert studio, from a list of expert studios providing expertise corresponding to the identified location associated with the customer's request, is matched up with the request. Thereafter, a primary connection, supporting an interactive videoconference session, is established between a first networked node associated with the customer and a second networked node associated with the designated expert studio.