Abstract: Certain aspects of the present disclosure support a local competitive learning rule applied in a computational network that leads to sparse connectivity among processing units of the network. The present disclosure provides a modification to the Oja learning rule, modifying the constraint on the sum of squared weights in the Oja rule. This constraining can be intrinsic and local as opposed to the commonly used multiplicative and subtractive normalizations, which are explicit and require the knowledge of all input weights of a processing unit to update each one of them individually. The presented rule provides convergence to a weight vector that is sparser (i.e., has more zero elements) than the weight vector learned by the original Oja rule. Such sparse connectivity can lead to a higher selectivity of processing units to specific features, and it may require less memory to store the network configuration and less energy to operate it.

Abstract: Embodiments of the present invention are directed to a matching method that implements a machine learning algorithm to serve one or more matches to an individual and to ensure that an outcome of each match is more likely than not to be successful. A recommendation engine uses the machine learning algorithm to perform predictive analysis to thereby select one or more members of an online community to match or pair with the individual, based on at least structured data, unstructured data or both of the individual and/or each selected member. The recommendation engine is configured to continuously learn from past user behavior, including the individual's, to further improve future matches provided to the individual.

Abstract: A simple format is disclosed and referred to as Elementary Network Description (END). The format can fully describe a large-scale neuronal model and embodiments of software or hardware engines to simulate such a model efficiently. The architecture of such neuromorphic engines is optimal for high-performance parallel processing of spiking networks with spike-timing dependent plasticity. Neuronal network and methods for operating neuronal networks comprise a plurality of units, where each unit has a memory and a plurality of doublets, each doublet being connected to a pair of the plurality of units. Execution of unit update rules for the plurality of units is order-independent and execution of doublet event rules for the plurality of doublets is order-independent.

Abstract: The present techniques extract attribute data of one or more classified members for one or more user attributes. With respect to a particular user attribute of the one or more user attributes, the present techniques determine initial attribute data intervals corresponding to the particular user attribute based on attribute data and classes of the classified members from the extracted attribute data. With respect to a classified member whose attribute data is missing for the particular user attribute, the present techniques set the attribute data as a preset missing value. The present techniques then merge the preset missing value into each of the initial user attribute data intervals and calculate a Maximum Posteriori Probability (MAP) Bayes estimate value respectively, and determine initial user attribute data intervals with a smallest MAP Bayes estimated value as final attribute data intervals corresponding to the particular user attribute.

Abstract: The problem of providing an efficient physical implementation of a (first) classifier defined by a first rule set, at least a part of which first classifier having a sparse distribution in Boolean space, is solved by (1) converting the first classifier, having a corresponding Boolean space, into a second classifier, wherein the second classifier has a corresponding Boolean space which is not semantically equivalent to the Boolean space corresponding to the first classifier, and wherein the second classifier is defined by a second set of rules which is smaller than the first set of rules defining the first classifier; and (2) defining a bit string transformation which transforms a first bit string into a second bit string, wherein applying the first bit string to the first classifier is equivalent to applying the second bit string to the second classifier. In at least some example embodiments, the first bit string includes packet header information.

Abstract: A classifier learning module trains video classifiers associated with a stored set of concepts derived from textual metadata of a plurality of videos. Specifically, a first type of classifier (e.g., a content-based classifier) and a second type of classifier (e.g., a text-based classifier) are trained, the classifiers when applied to a video indicating a likelihood that the video represents one or more concepts corresponding to the classifier. The first type of classifier can be used to determine the training set for the second type of classifier. The learning process does not require any concepts to be known a priori, nor does it require a training set of videos having training labels manually applied by human experts. Scores from the first type of classifier are combined with scores from the second type of classifier to obtain video classification of enhanced accuracy.

Abstract: Adaptive proportional-integral-derivative controller apparatus of a plant may be implemented. The controller may comprise an encoder block utilizing basis function kernel expansion technique to encode an arbitrary combination of inputs into spike output. The basis function kernel may comprise one or more operators configured to manipulate basis components. The controller may comprise spiking neuron network operable according to reinforcement learning process. The network may receive the encoder output via a plurality of plastic connections. The process may be configured to adaptively modify connection weights in order to maximize process performance, associated with a target outcome. Features of the input may be identified and used for enabling the controlled plant to achieve the target outcome.

Abstract: Aspects of the subject matter described herein relate to predicting and using search engine switching behavior. In aspects, switching components receive a representation of user interactions with at least one browser. The switching components derive information from the representation that is useful in predicting whether a user will switch search engines. The derived information and information about a user's current interaction with a browser is then used by a switch predictor to predict whether the user will switch search engines. This prediction may be used in a variety of ways examples of which are given herein.

Abstract: Techniques to estimate the probability of a future event occurring are described. The techniques include decomposing a data input stream to build a database of precursor data and building predictive models using the precursor data. Also disclosed are techniques in which by using a search engine to search a database of models to find a model and a user can query a found model to develop an inference of the likelihood of the future event.

Abstract: An analysis tool for causing a computer to use information gain of attributes and a classification algorithm to classify new records in a set of data by taking into account the predictive value of the attributes and the effect of the new record.

Abstract: A system, method and computer program product for scalable, rule-based processing, including an automaton builder for constructing automatons corresponding to trigger rules and word lists that are employed by the trigger rules, and a rule evaluator for evaluating any of the trigger rules with respect to an input document by selecting any of the automatons to evaluate a given one of the trigger rules, parsing the input document using the selected automatons, determining whether conditions of the given trigger rule are met, and identifying any actions that are associated with the given trigger rule.

Abstract: A system, method and apparatus for automatically adapting power grid usage by controlling internal and/or external power-related assets of one or more users in response to power regulation and/or frequency regulation functions in a manner beneficial to both the power grid itself and the users of the power grid.

Abstract: An apparatus, device, methods, computer program product, and system are described that provide a graphical illustration of a first possible outcome of a use of a treatment parameter with respect to at least one body portion, based on a first dataset associated with a first predictive basis, and that modify the graphical illustration to illustrate a second possible outcome of the use of the treatment parameter, based on a second dataset associated with a second predictive basis.

Abstract: The present invention 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. The method of the present invention comprises 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: An automatic updating apparatus includes a traffic receiver that receives numbers per unit time of the access of more than one menu displayed in a screen and calculates rates of variability with respect to the numbers of the access to each menu, and a menu updating unit that updates a menu display in the screen based on the rates of variability.

Abstract: Embodiments relate to classification of transactions based upon analysis of multiple variables. For a purchase transaction, such variables can include but are not limited to: buying location, source system, line of business, cost center, functional area, supplier capabilities, item description, account description, organization, department, custom parameters, and others. Embodiments may rely upon one or more classification schemes, such as statistical classification, semantic classification, and/or knowledge base classification, taken alone or in combination. In a purchase transaction, classification based on multivariate analysis facilitates identification of a purchased item or service, and hence accuracy in classifying and assigning a central classification code. Particular embodiments may include a feature allowing user review/revision of category assignments via a feedback loop linked to past classification.

Abstract: A method of adapting an event processing component. The method comprises designating an event processing component having a plurality of event processing agents which carry out a plurality of rules to process a plurality of events, selecting at least one rules correctness requirement, and automatically adjusting, using a processor, the plurality of event processing rules to comply with the at least one correctness requirement.

Abstract: A method for learning connections between nonlinear oscillators in a neural network comprising the steps of providing a plurality of nonlinear oscillators, with each respective oscillator producing an oscillation distinct from the others in response to an input and detecting an input at an at least first oscillator of the plurality of nonlinear oscillators. Detecting an input at an at least a second oscillator of the plurality of nonlinear oscillators, comparing the oscillation of the at least first oscillator to the oscillation of the at least second oscillator at a point in time, and determining whether there is coherency between the oscillation of the at least first oscillator and the oscillation of the at least second oscillator. Changing at least one of the amplitude and phase of a connection between the at least first oscillator and the at least second least oscillator as a function coherency between the at least first oscillator and the oscillation of the at least second oscillator.

Abstract: Multi-task regression or classification includes optimizing parameters of a Bayesian model representing relationships between D features and P tasks, where D?1 and P?1, respective to training data comprising sets of values for the D features annotated with values for the P tasks. The Bayesian model includes a matrix-variate prior having features and tasks dimensions of dimensionality D and P respectively. The matrix-variate prior is partitioned into a plurality of blocks, and the optimizing of parameters of the Bayesian model includes inferring prior distributions for the blocks of the matrix-variate prior that induce sparseness of the plurality of blocks. Values of the P tasks are predicted for a set of input values for the D features using the optimized Bayesian model. The optimizing also includes decomposing the matrix-variate prior into a product of matrices including a matrix of reduced rank in the tasks dimension that encodes correlations between tasks.

Abstract: A method of routing data through a router in a communications network, the method comprising receiving one or more data packets, each packet having a respective destination address and applying a lookup algorithm to each packet, said lookup algorithm being arranged to determine a respective route along which each packet is to be transmitted towards its destination address by searching an associated hierarchical data structure containing routing information for each packet. The method comprising forwarding each packet for transmission to its respective destination address, wherein said lookup algorithm comprises an adaptive learning component that is configured to dynamically identify an optimum starting position for searching within said hierarchical data structure, for each of the data packets, based on the results of one or more earlier searches.