Abstract: Embodiments of the present invention relate to meeting latency constraints in a multi-core neurosynaptic network. In one embodiment of the present invention, a method of and computer program product for power-driven synthesis under latency constraints is provided. Power consumption of a neurosynaptic network is modeled as wire length. The neurosynaptic network comprises a plurality of neurosynaptic cores. Each of the plurality of neurosynaptic cores is modeled as a node in a placement graph. The graph has a plurality of edges. A weight is assigned to each of the plurality of edges based on a spike frequency. An arrangement of the neurosynaptic cores is determined. The arrangement comprises a length of each of the plurality of edges. A maximum length is compared to the length of each of the plurality of edges. The weight of at least one of the plurality of edges is increased where the length is greater than the maximum length.

Abstract: Automatically create abstractions of large sets of data and then probabilistic inferences based on the abstractions. The probabilistic inference is derived from the logical hierarchy using Bayesian statistics to infer a probabilistic event based upon a characteristic of the data in a hierarchy of synthetic events. The logical hierarchy of a set of a plurality of synthetic events is related by at least one characteristic of data is built by accessing a first set of data. The first set of data is organized based on a first characteristic. A second set of data different than the first set of data is accessed. A second set of data based is organized based on a second characteristic. The first characteristic and the second characteristic are processed to generate a synthetic event. The synthetic event is a third set of data representing a result of a mathematical computation defined by an operation S(p1)==>F(p2).

Abstract: Methods and systems for training a neural network are provided. One system includes one or more components executed by one or more computer subsystems. The one or more components include a neural network configured for determining inverted features of input images in a training set for a specimen input to the neural network, a forward physical model configured for reconstructing the input images from the inverted features thereby generating a set of output images corresponding to the input images in the training set, and a residue layer configured for determining differences between the input images in the training set and their corresponding output images in the set. The one or more computer subsystems are configured for altering one or more parameters of the neural network based on the determined differences thereby training the neural network.

Abstract: A machine learning apparatus, which learns a condition associated with a filter unit for filtering an analog input signal, includes a state observer for observing a state variable that includes at least one of a noise component and noise amount of an output signal from the filter unit and a responsivity to the input signal; and a learner for learning the condition associated with the filter unit in accordance with a training data set that includes the state variable.

Abstract: A method and system are presented of automatically suggesting rules for data stored in a table, with the table comprising a plurality of columns. The table is profiled to identify a content type for each of one or more of the plurality of columns. A rule knowledge base is accessed to locate rules specified for identified content types. Then, one or more of the located rules specified for identified content types are presented as suggestions. Acceptance of one or more of the suggested rules is received from a user, and the received validations are stored in the rule knowledge base. The accepted rules are applied to data for quality detection and monitoring. Embodiments are also described where columns are suggested based on a given rule.

Abstract: The Support Vector Machine (SVM) has been used in a wide variety of classification problems. The original SVM uses the hinge loss function, which is nondifferentiable and makes the problem difficult to solve in particular for regularized SVMs, such as with l1-norm. The Huberized SVM (HSVM) is considered, which uses a differentiable approximation of the hinge loss function. The Proximal Gradient (PG) method is used to solving binary-class HSVM (BHSVM) and then generalized to multi-class HSVM (MHSVM). Under strong convexity assumptions, the algorithm converges linearly. A finite convergence result about the support of the solution is given, based on which the algorithm is further accelerated by a two-stage method.

Abstract: This disclosure relates generally to multi-sensor data, and more particularly to summarizing multi-sensor data. In one embodiment, the method includes computing plurality of histograms from sensor data associated with a plurality of sensors. The respective histograms of each sensor are clustered into a first plurality of sensor-clusters, and a first set of rules is extracted therefrom. First set of rules defines patterns of histograms of a set of sensors occurring frequently over a time-period. Two or more sensor-clusters from amongst the first plurality of sensor-clusters are merged selectively to obtain a second plurality of sensor-clusters. Second set of rules are extracted from the second plurality of sensor-clusters, and a set of correlated sensors are identified therefrom based on the second set of rules. Third set of rules are extracted from the set of correlated sensors, the third set of rules summarizes the multi-sensor data to represent prominent co-occurring sensor behaviors.

Abstract: The invention relates to predictive browsing. A set of words for use with an experience matrix are formed, wherein the words are descriptive of a context of a system such as a current web page, and wherein said experience matrix comprises sparse vectors associated with words. At least a part of at least one sparse vector of said experience matrix is accessed to form a prediction output, and suggestions of web pages are provided to a user in response to said prediction output.

Abstract: At least one label prediction model is trained, or learned, using training data that may comprise training instances that may be missing one or more labels. The at least one label prediction model may be used in identifying a content item's ground-truth label set comprising an indicator for each label in the label set indicating whether or not the label is applicable to the content item.

Abstract: An application analysis computer obtains reports from user terminals identifying operational states of instances of an application being processed by the user terminals. Sequences of the operational states that the instances of the application have transitioned through while being processed by the user terminals are identified. Common operational states that occur in a plurality of the sequences are identified. For each of the common operational states, a frequency of occurrence of the common operational state is determined. For each state transition between the common operational states in the sequences, a frequency of occurrence of the state transition is determined. State predictive metrics are generated based on the frequencies of occurrence of the common operational states and the frequencies of occurrence of the state transitions. The state predictive metrics are communicated, such as to an application server to control access to the application by user terminals.

Abstract: A resistive memory cell is connected in circuitry which has a first input terminal for applying neuron input signals including a read portion and a write portion. The circuitry includes a read circuit producing a read signal dependent on resistance of the memory cell, and an output terminal providing a neuron output signal, dependent on the read signal in a first state of the memory cell. The circuitry also includes a storage circuit storing a measurement signal dependent on the read signal, and a switch set operable to supply the read signal to the storage circuit during application of the read portion of each neuron input signal to the memory cell, and, after application of the read portion, to apply the measurement signal in the apparatus to enable resetting of the memory cell to a second state.

Abstract: Mechanisms for evaluating an evidential statement in a corpus of evidence are provided. An evidential statement is received for determining a level of confidence in a hypothetical ontological link of an ontology. A source of the evidential statement is identified and a grading of the source of the evidential statement is determined based on a source grading measurement value indicative of a degree of reliability and credibility of the source. An indication of trustworthiness of the evidential statement is generated based on the source grading measurement value. A representation of the indication of trustworthiness of the evidential statement is output in association with the evidential statement.

Abstract: Disclosed is a system for determining an expert of one or more subjects on a web-based platform. The system comprises a mining module for mining activity data of at least one user of a plurality of users from the web-based platform. The mining module may further compare the activity data with one or more subjects. The mining module may further label the activity data to a subject of the one or more subjects. A scoring module may assign performance points to the at least one user associated to the activity data. The scoring module may further assign subject points to the subject. The scoring module may further generate an activity gauge for the at least one user based on the performance points assigned and the subject points. The scoring module may further classify the at least one user as the expert of the subject.

Abstract: An indication of a data source to be used to train a linear prediction model is obtained. The model is to generate predictions using respective parameters assigned to a plurality of features derived from observation records of the data source. The parameter values are stored in a parameter vector. During a particular learning iteration of the training phase of the model, one or more features for which parameters are to be added to the parameter vector are identified. In response to a triggering condition, parameters for one or more features are removed from the parameter vector based on an analysis of relative contributions of the features represented in the parameter vector to the model's predictions. After the parameters are removed, at least one parameter is added to the parameter vector.

Abstract: A system is provided for event-based monitoring of a subject's well-being within an unattended setting. A plurality of sensors are disposed within the setting for sensing disparate events, and an analytics processing portion is coupled to the sensors to collectively acquire sensing data therefrom, and map a plurality of sensed data points for a selected combination of disparate events to a conduct adaptively characterized for the subject. The mapping occurs according to a set of pre-established reference event patterns, relative to which each characterized conduct is screened for excessive aberration. The analytics processing portion actuates generation of a graphic user interface displaying at least one reporting page. The reporting page contains for each characterized conduct certain graphic indicia determined responsive to the screening thereof.

Abstract: User interactions are categorized into predefined hierarchical categories by classifying user interactions, such as queries, during a user interaction session by labeling text data into predefined hierarchical categories, and building a scoring model. The scoring model is then executed on untagged user interaction data to classify the user interactions into either action-based or information-based interactions.

Abstract: Aspects of the present invention disclose a method, computer program product, and system for generating target text based on target data. The method includes one or more processors decomposing one or more portions of text into at least one corresponding keyword and at least one corresponding template. The method further includes learning a classification model associated with selecting a template based on a category of a keyword. The method further includes identifying a target keyword that is represented by target data. The method further includes selecting a target template that is used to represent the target data based on a category associated with the identified target keyword utilizing the classification model. The method further includes generating target text that represents the target data based on the selected text template based on the selected target template and the identified target keyword.

Abstract: A method and system are provided. The method includes performing, by a logs-to-time-series converter, a logs-to-time-series conversion by transforming a plurality of heterogeneous logs into a set of time series. Each of the heterogeneous logs includes a time stamp and text portion with one or more fields. The method further includes performing, by a time-series-to-sequential-pattern converter, a time-series-to-sequential-pattern conversion by mining invariant relationships between the set of time series, and discovering sequential message patterns and association rules in the plurality of heterogeneous logs using the invariant relationships. The method also includes executing, by a processor, a set of log management applications, based on the sequential message patterns and the association rules.

Abstract: Aspects of the present invention disclose a method, computer program product, and system for generating target text based on target data. The method includes one or more processors decomposing one or more portions of text into at least one corresponding keyword and at least one corresponding template. The method further includes learning a classification model associated with selecting a template based on a category of a keyword. The method further includes identifying a target keyword that is represented by target data. The method further includes selecting a target template that is used to represent the target data based on a category associated with the identified target keyword utilizing the classification model. The method further includes generating target text that represents the target data based on the selected text template based on the selected target template and the identified target keyword.

Abstract: Methods and apparatus are provided for identifying environmental stimuli in an artificial nervous system using both spiking onset and spike counting. One example method of operating an artificial nervous system generally includes receiving a stimulus; generating, at an artificial neuron, a spike train of two or more spikes based at least in part on the stimulus; identifying the stimulus based at least in part on an onset of the spike train; and checking the identified stimulus based at least in part on a rate of the spikes in the spike train. In this manner, certain aspects of the present disclosure may respond with short response latencies and may also maintain accuracy by allowing for error correction.