Abstract: A system and method a Multi-Task Multi-View (M2TV) learning problem. The method uses the label information from related tasks to make up for the lack of labeled data in a single task. The method further uses the consistency among different views to improve the performance. It is tailored for the above complicated dual heterogeneous problems where multiple related tasks have both shared and task-specific views (features), since it makes full use of the available information.

Abstract: A method of performing constraint satisfaction problem (CSP) solving. The method comprises evaluating the nodes of a CSP tree for a TRUE or FALSE state according to a script, wherein the tree models a CSP problem and the script associates rules and corresponding actions with each tree node. During a step of evaluating a specific node, there is dynamically computed a backjump pointer according to a specific rule associated with the specific node, wherein the value of the computed backjump pointer identifies an ancestor node of the tree. Upon a failure to evaluate the node to a TRUE state, the evaluation is restarted at the ancestor node identified by the backjump pointer.

Abstract: A cut optimization system controls chatter in a machine tool during a cutting operation. A microphone is configured to capture acoustic noise emitted by the machine tool during the cutting operation and to generate an AC signal corresponding to the captured acoustic noise. A filter is configured to attenuate frequencies of the AC signal outside of a frequency band and a rectifier is configured to rectify the filtered AC signal into a DC component. A controller is configured to compare the DC component with a threshold value and, if the DC component is greater than the threshold, cyclically vary the rotational speed of a spindle in the machine tool from a commanded speed.

Abstract: A device (100) for the quality analysis of system models (10, 12) has a testing tool (1) that is designed to receive a plurality of system models (10, 12) from a plurality of modeling tools (11, 13), convert it into a unified data model (2) and check it for quality with the aid of unified quality rules (3).

Abstract: An information processing apparatus creates, for each of a plurality of nodes, a query to be executed for a learning pattern input to the node; inputs a plurality of learning patterns to a root node of the plurality of nodes; executes, for the learning pattern input to each node, the query created for the node; determines whether the query has been effectively executed for the individual learning pattern input to each node; distributes and inputs, to a lower node of each node, an individual learning pattern for which it has been determined in the determining that the query was effectively executed in the node; deletes a learning pattern for which it has been determined in the determining that the query was not effectively executed in each node; and stores an attribute of the learning pattern input to a terminal node of the plurality of nodes in association with the node.

Abstract: Both a cascade and a multichannel joint Bayesian estimator are provided for suppressing acoustic echo. An expansion basis (Power/Fourier series) is selected to convert a sample-based input signal xt into a DFT-domain multichannel signal [X?,1, . . . X?,p]. The posterior of unknown states (e.g., mean ?? and covariance P? of the echo path W? and the mean â? and covariance Q? of the nonlinear coefficients a?; or channel-wise mean ??,i and multichannel covariance P? of a compound quantity formed by merging together the echo path W? and the ith nonlinear coefficient a?,i) and model parameters ?? are estimated; and Kalman gain factor(s) K? are computed for optimal adaptation of the posterior of unknown states. An echo signal ?? is estimated using the multichannel input signal [X?,1, . . . X?,p] and the adapted posterior; and an error signal E? is generated.

Abstract: Methods, systems, and media are provided for a dynamic batch strategy utilized in parallelization of online learning algorithms. The dynamic batch strategy provides a merge function on the basis of a threshold level difference between the original model state and an updated model state, rather than according to a constant or pre-determined batch size. The merging includes reading a batch of incoming streaming data, retrieving any missing model beliefs from partner processors, and training on the batch of incoming streaming data. The steps of reading, retrieving, and training are repeated until the measured difference in states exceeds a set threshold level. The measured differences which exceed the threshold level are merged for each of the plurality of processors according to attributes. The merged differences which exceed the threshold level are combined with the original partial model states to obtain an updated global model state.

Abstract: A plan representing a final solution to a problem is obtained efficiently, if the problem changes while being solved, by identifying the solution elements not affected by the change, and by reusing those solution elements.

Abstract: A “Compatibility Tester” extracts observable behaviors from different system implementations (e.g., software, operating systems, device drivers, hardware interfaces, etc.), then performs compatibility testing via comparisons of the extracted behaviors. Traces (e.g., bus level signals between controllers and hardware devices or commands between software components) representing observable behaviors of different implementations of a system are captured. Temporal and structural rules are then mined from these traces. The mined rules (or a model constructed from those rules) are used by a “Rule Checking” process that determines whether reference rules (mined from a known compatible system) are compatible with rules mined from test traces of a second system. Invalid rules are flagged as behavioral compatibility bugs.

Abstract: A method of ocean property forecasting. The method includes generating one of a magnetic field prediction and an electric field prediction between a first observation point and a plurality of uniform ocean cells by generating a vector sum of magnetic field contributions from a plurality of magnetic field vectors and a vector sum of electric field contributions from a plurality of electric field vectors, respectively. The method further includes generating one of a magnetic field forecast and an electric field forecast of the ocean volume at a second observation point in the ocean volume by re-using the magnetic field vectors and electric field vectors and scaling one of the plurality of magnetic field vectors and the plurality of electric field vectors by a plurality of tri-axial equivalent electrical currents with identical geometry to the first observation point.

Abstract: Methods and systems are described for simplifying a causal influence model that describes influence of parent nodes Xi (i=1, . . . , n) on possible states of the child node Y. The child node Y and each one of the parent nodes Xi (i=1, . . . , n) are assumed to be either a discrete Boolean node having states true and false, a discrete Ordinal node having a plurality of ordered states; and a Categorical node having a plurality of unordered states. The influence of each parent node Xi on the child node Y is assumed to be a promoting influence and an inhibiting influence. User interfaces are described that incorporate these specific node types.

Abstract: The invention relates generally to systems, devices and methods for the efficient use of utilities, more particularly to the distribution and provision of electricity supply at appropriate voltages, monitoring and usage by end devices, and to facilitating consumers in changing their energy usage behavior, and to adopt and easily install appropriate sustainable, energy efficient or renewable technologies. Said end devices typically including traditional electric, electronic and lighting appliances requiring AC or DC power provision or low voltage DC power via AC/DC converters.

Abstract: In one embodiment, a cognitive radio dynamic spectrum access sensing system is configured to incorporate a hidden Markov model and a risk function to determine the primary user state sequence with the minimum amount of associated risk. The system comprises one or more weighted cost factors that can be used to handle a missed detection sensing error differently from a false alarm sensing error. The system further comprises a complete forward partial backward computation designed to increase sensing accuracy with limited effect on sensing delay and complexity.

Abstract: Computer-implemented methods, systems, and articles of manufacture for determining the importance of a data item. A method includes: (a) receiving a node graph; (b) approximating a number of neighbor nodes of a node; and (c) calculating a average shortest path length of the node to the remaining nodes using the approximation step, where this calculation demonstrates the importance of a data item represented by the node. Another method includes: (a) receiving a node graph; (b) building a decomposed line graph of the node graph; (c) calculating stationary probabilities of incident edges of a node graph node in the decomposed line graph, and (d) calculating a summation of the stationary probabilities of the incident edges associated with the node, where the summation demonstrates the importance of a data item represented by the node. Both methods have at least one step carried out using a computer device.

Abstract: Methods and systems for creating ensembles of hypersurfaces in high-dimensional feature spaces, and to machines and systems relating thereto. More specifically, exemplary aspects of the invention relate to methods and systems for generating supervised hypersurfaces based on user domain expertise, machine learning techniques, or other supervised learning techniques. These supervised hypersurfaces may optionally be combined with unsupervised hypersurfaces derived from unsupervised learning techniques. Lower-dimensional subspaces may be determined by the methods and systems for creating ensembles of hypersurfaces in high-dimensional feature spaces. Data may then be projected onto the lower-dimensional subspaces for use, e.g., in further data discovery, visualization for display, or database access.

Abstract: A control system includes a control device for controlling a plurality of actuators by executing a plurality of functions based sensor values from a plurality of external sensors and a function providing device for providing the control device with the functions to be executed by the control device. The function providing device accepts a program including a plurality of functions from a user device, finds a new function among the functions which has not been received yet and pass the same function to the control device. The control device executes functions passed from the function providing device based on values of the external sensors and controls the actuators by transmitting values of output arguments of the functions to the actuators.

Abstract: A method and system are provided for encapsulation and re-use of a model. A modeling platform creates a governance contract governing the exportation and modification of elements of the model by other models. In response to a request to import at least one element of the model into a second model, the modeling platform exports the requested at least one element by reference only if the element has a governance setting that allows for exportation. In response to a request to modify at least one imported element of the model in the second model, the modeling platform determines the governance setting for the element. Then, the modeling platform modifies the element by composing the modification in the second model only if the element has a governance setting that allows modification.

Abstract: A driving maneuver assisting apparatus includes a learning section configured to learn a driving-behavior pattern of a driver for a predetermined duration; a non-steady-state degree calculating section configured to calculate a non-steady-state degree by comparing a current driving-behavior pattern with the driving-behavior pattern learned by the learning section, wherein the non-steady-state degree represents how different the current driving-behavior pattern is from the driving-behavior pattern learned by the learning section; a learning level calculating section configured to calculate a learning level of the learning section; and a notifying section configured to notify the driver of a maneuver assisting information for inducing the driving-behavior pattern learned by the learning section in accordance with the learning level calculated by the learning level calculating section, when the non-steady-state degree calculated by the non-steady-state degree calculating section exceeds a threshold value.

Abstract: Computer-implemented systems and methods are provided for generating a data model. A variable predictiveness determination is performed on the population of candidate variables. A plurality of variables from the population of candidate variables are selected as a selected set based on the variable predictiveness values. A plurality derived variables are generated based on variables in the rejected set without consideration of any variables in the selected set. One or more derived variables are selected as based on derived variable predictiveness values of the derived variables, and the selected set and the one or more selected derived variables are stored as the model input variables for the data model.

Abstract: A policy scheduler scheduling a policy is provided. The policy scheduler receives the policy for a system and information of a current state of the system. The policy scheduler evaluates one or more rules based on the current state of the system and generates a new rule via an evolutionary algorithm based on the information of the current state of the system. The policy scheduler adds the newly generated rule into the one or more rules and schedules the policy based on the one or more rules including the newly generated rule.