An efficient neural network computing technique capable of synthesizing two sets of output signal data from a single input signal data set. The method and device of the invention involves a unique integration of autoassociative and heteroassociative neural network mappings, the autoassociative neural network mapping enabling a quality metric for assessing the generalization or prediction accuracy of the heteroassociative neural network mapping.

According to one aspect of the present invention, a method is provided in which a search request containing search criteria is received from a first user. The search criteria include at least one search category. Filter search specification associated with the at least one search category is determined. The filter search specification is used to specify a particular type of records in the at least one search criteria to be selected for the search request. A search is performed based upon the fil...

In one embodiment, the invention includes sorting a set of constraints based on weights associated with each constraint, the constraints being capable of being satisfied based on values selected for corresponding associated parameters, selecting a constraint that is unsatisfied and that has the greatest weight, selecting parameters associated with the constraint and finding parameter values to satisfy the constraint, after satisfying the selected unsatisfied constraint, selecting a further unsat...

A method for analyzing biological data includes classifying a first set of biological data in a first classifier, classifying a second set of biological data in a second classifier, combining the results of the first classifier with the results of the second classifier, and analyzing the results as a function of the similarity measure of the first classifier and the similarity measure of the second classifier.

Decision-support systems, such as neural networks, for assessing the risk of delivery within a predetermined time or the likelihood of preterm delivery are provided.

A system for tracking an activity in a healthcare environment includes a master station having a processor, a memory, and a transceiver, badges coupled to objects, each transmitting a unique badge ID, and sensors disposed throughout the environment, each including a transceiver for receiving badge IDs and transmitting to the master station transceiver the badge IDs and a unique sensor ID that relates the sensor to a location. The master station processor determines object locations from the badg...

An apparatus, a carrier medium carrying computer readable code to implement a method, and a method for searching for a plurality of patterns definable by complex expressions, and further, for efficiently generating data for such searching. One method includes accepting or determining a plurality of state machines for searching for a plurality of patterns, merging the state machines to form a merged state machine, and storing a data structure describing the merged state machine, including state d...

A hyper-spectral analysis method for characterizing or distinguishing diverse elements within hyper-spectral images. A plurality of patches of pixels from within the hyper-spectral images are extracted as being patches around pixels of the elements to be characterized or distinguished. The statistics of spectra for each patch of pixels are computed. A first classifier is computed from frequency-wise standard deviation of the spectra in each patch and a set of second classifiers are computed from...

A method and apparatus for retraining skew compensation in an interface is presented. In one embodiment, a retraining interval is determined, and counters in the transmitting agent and receiving agent count up until the retraining interval is reached. A tracking unit used to select one of several interpolated clocks may then be powered up, and a special retraining phit may be sent across the interface. During the retraining process, the transfer of flits into and out of the flow-control mechanis...

A system and method for fault detection is provided. The fault detection system provides the ability to detect symptoms of fault in turbine engines and other mechanical systems that have nonlinear relationships between two or more variables. The fault detection system uses a neural network to perform feature extraction from data for representation of faulty or normal conditions. The values of extracted features, referred to herein as scores, are then used to determine the likelihood of fault in ...