1137 results about "Near neighbor" patented technology

Predictive modeling of consumer financial behavior, including determination of likely responses to particular marketing efforts, is provided by application of consumer transaction data to predictive models associated with merchant segments. The merchant segments are derived from the consumer transaction data based on co-occurrences of merchants in sequences of transactions. Merchant vectors represent specific merchants, and are aligned in a vector space as a function of the degree to which the merchants co-occur more or less frequently than expected. Consumer vectors are developed within the vector space, to represent interests of particular consumers by virtue of relative vector positions of consumer and merchant vectors. Various techniques, including clustering, supervised segmentation, and nearest-neighbor analysis, are applied separately or in combination to generate improved predictions of consumer behavior.

A method and apparatus for processing an audio sound source to create four-dimensional spatialized sound. A virtual sound source may be moved along a path in three-dimensional space over a specified time period to achieve four-dimensional sound localization. A binaural filter for a desired spatial point is applied to the audio waveform to yield a spatialized waveform that, when the spatialized waveform is played from a pair of speakers, the sound appears to emanate from the chosen spatial point instead of the speakers. A binaural filter for a spatial point is simulated by interpolating nearest neighbor binaural filters chosen from a plurality of pre-defined binaural filters. The audio waveform may be processed digitally in overlapping blocks of data using a Short-Time Fourier transform. The localized sound may be further processed for Doppler shift and room simulation.

A system and process for identifying corresponding points among multiple images of a scene is presented. This involves a multi-view matching framework based on a new class of invariant features. Features are located at Harris corners in scale-space and oriented using a blurred local gradient. This defines a similarity invariant frame in which to sample a feature descriptor. The descriptor actually formed is a bias/gain normalized patch of intensity values. Matching is achieved using a fast nearest neighbor procedure that uses indexing on low frequency Haar wavelet coefficients. A simple 6 parameter model for patch matching is employed, and the noise statistics are analyzed for correct and incorrect matches. This leads to a simple match verification procedure based on a per feature outlier distance.

A method of automatically labeling of unlabeled text data can be practiced independent of human intervention, but that does not preclude manual intervention. The method can be used to extract relevant features of unlabeled text data for a keyword search. The method of automated labeling of unlabeled text data uses a document collection as a reference answer set. Members of the answer set are converted to vectors representing centroids of unknown groups of unlabeled text data. Unlabeled text data are clustered relative to the centroids by a nearest neighbor algorithm and the ID of the relevant answer is assigned to all documents in the cluster. At this point in the process, a supervised machine learning algorithm is trained on labeled data, and a classifier for assigning labels to new text data is output. Alternatively, a feature extraction algorithm may be run on classes generated by the step of clustering, and search features output which index the unlabeled text data.

An apparatus and method for obtaining facies of geological formations for identifying mineral deposits is disclosed. Logging instruments are moved in a bore hole to produce log measurements at successive levels of the bore hole. The set of measurements at each such level of the bore hole interval is associated with reference sample points within a multidimensional space. The multidimensional scatter of sample points thus obtained is analyzed to determine a set of characteristic modes. The sample points associated with characteristic modes are grouped to identify clusters. A facies is designated for each of the clusters and a graphic representation of the succession of facies as a function of the depth is thus obtained. To identify the clusters, a "neighboring index" of each log measurement point in the data set is calculated. Next, small natural groups of points are formed based on the use of the neighboring index to determine a K-Nearest-Neighbor (KNN) attraction for each point. Independently of the natural group formation, an optimal number of clusters is calculated based on a Kernel Representative Index (KRI) and based on a user-specified resolution. Lastly, based on the data calculated from the prior steps, final clusters are formed by merging the smaller clusters.

A magnetoresistive element comprises an exchange coupling film having a under layer, an antiferromagnetic film and a ferromagnetic film, which are laminated in that order, the under layer including a metal having a face centered cubic crystal structure or hexagonal closest packing crystal structure which have a longer nearest neighbor atomic distance than that of the antiferromagnetic film. With this construction, it is possible to improve the exchange coupling field and to satisfy a stable output over a long period of time. A magnetoresistive element having a dual spin valve structure has a magnetization adjusting layer, which is antiferromagnetically connected to a pinned layer via an anti-parallel connection layer, to adjust the value of the product of the saturation magnetization of each of the magnetization adjusting layer and the pinned layer by the thickness thereof. Moreover, a magnetoresistance head use a giant magnetoresistance effect, and has at least one pair of pinned layer and free layer arranged via a non-magnetic spacer layer. The pinned layer has a pair of ferromagnetic layers which have different compositions and different coercive forces and which are antiferromagnetically connected to each other via a connection layer, so that the effective exchange coupling field of the pinned layer is 200 Oe or more.

A semiconductor device having a semiconductor layer, includes: a first impurity atom having a covalent bond radius larger than a minimum radius of a covalent bond of a semiconductor constituent atom of a semiconductor layer; and a second impurity atom having a covalent bond radius smaller than a maximum radius of the covalent bond of the semiconductor constituent atom; wherein the first and second impurity atoms are arranged in a nearest neighbor lattice site location and at least one of the first and second impurity atoms is electrically active.

The invention discloses a personalized commodity recommending method which integrates attributes and structural similarity. In the method, users and commodities are used as nodes with characteristic information to be mapped to a network by integrating the attribute information and structural similarity information, and an information network chart is established according to the purchasing relation between customers and the commodities; and interests and preference among user node pairs are measured by the integrated attributes and structural similarity in the information network chart, and the nearest neighbor is selected by the interests and the preference to improve the accuracy of recommending. On the basis of the recommending method, the invention also discloses a personalized commodity recommending method which integrates the measurement of the attributes and the structural similarity. In the system, the interests and the preference of the users are measured accurately by a computing method of integrating the similarity of the attributes and the similarity of node structure backgrounds in the information network chart, and the generation efficiency of the nearest neighbor is improved by utilizing clustering technology. The method and the system can be applied to electronic commerce, and provide personalized commodity recommending for the users.

A general computer-implement method and apparatus to optimize problem layout on a massively parallel supercomputer is described. The method takes as input the communication matrix of an arbitrary problem in the form of an array whose entries C(i, j) are the amount to data communicated from domain i to domain j. Given C(i, j), first implement a heuristic map is implemented which attempts sequentially to map a domain and its communications neighbors either to the same supercomputer node or to near-neighbor nodes on the supercomputer torus while keeping the number of domains mapped to a supercomputer node constant (as much as possible). Next a Markov Chain of maps is generated from the initial map using Monte Carlo simulation with Free Energy (cost function) F=Σ_i,jC(i,j)H(i,j)—where H(i,j) is the smallest number of hops on the supercomputer torus between domain i and domain j. On the cases tested, found was that the method produces good mappings and has the potential to be used as a general layout optimization tool for parallel codes. At the moment, the serial code implemented to test the method is un-optimized so that computation time to find the optimum map can be several hours on a typical PC. For production implementation, good parallel code for our algorithm would be required which could itself be implemented on supercomputer.

A peer-to-peer storage system includes a storage coordinator that centrally manages distributed storage resources in accordance with system policies administered through a central administrative console. The storage resources, or “nodes,” are otherwise unused portions of storage media, e.g., hard disks, that are included in the devices such as personal computers, workstations, laptops, file servers, and so forth, that are connected to a corporate computer network, and are thus otherwise available only individually to the respective devices. The storage coordinator assigns the nodes to various “replication groups” and allocates the storage resources on each of the nodes in a given group to maintaining dynamically replicated versions of the group files. The storage nodes in a given group perform dynamic file replication and synchronization operations by communicating directly, that is, peer-to-peer, using a message-based protocol. The storage coordinator also manages distributed searches of file content on the network by selecting one node from each group to search through the associated group files. The selected nodes report the search results back to the storage coordinator, which organizes the results and provides them to the user. Thereafter, in response to a request for various files by the user, the storage coordinator instructs the nodes that are near neighbors of the user to provide the requested files. The storage coordinator thus ensures that the amount of the network bandwidth consumed by the search operation is minimized.

Characterization of carotid atherosclerosis and classification of plaque into symptomatic or asymptomatic along with the risk score estimation are key steps necessary for allowing the vascular surgeons to decide if the patient has to definitely undergo risky treatment procedures that are needed to unblock the stenosis. This application describes a statistical (a) Computer Aided Diagnostic (CAD) technique for symptomatic versus asymptomatic plaque automated classification of carotid ultrasound images and (b) presents a cardiovascular stroke risk score computation. We demonstrate this for longitudinal Ultrasound, CT, MR modalities and extendable to 3D carotid Ultrasound. The on-line system consists of Atherosclerotic Wall Region estimation using AtheroEdge™ for longitudinal Ultrasound or Athero-CTView™ for CT or Athero-MRView from MR. This greyscale Wall Region is then fed to a feature extraction processor which computes: (a) Higher Order Spectra; (b) Discrete Wavelet Transform (DWT); (c) Texture and (d) Wall Variability. The output of the Feature Processor is fed to the Classifier which is trained off-line from the Database of similar Atherosclerotic Wall Region images. The off-line Classifier is trained from the significant features from (a) Higher Order Spectra; (b) Discrete Wavelet Transform (DWT); (c) Texture and (d) Wall Variability, selected using t-test. Symptomatic ground truth information about the training patients is drawn from cross modality imaging such as CT or MR or 3D ultrasound in the form of 0 or 1. Support Vector Machine (SVM) supervised classifier of varying kernel functions is used off-line for training. The Atheromatic™ system is also demonstrated for Radial Basis Probabilistic Neural Network (RBPNN), or Nearest Neighbor (KNN) classifier or Decision Trees (DT) Classifier for symptomatic versus asymptomatic plaque automated classification. The obtained training parameters are then used to evaluate the test set. The system also yields the cardiovascular stroke risk score value on the basis of the four set of wall features.

Method and systems of traversing through a domain is provided. One method comprises getting a set of widely spaced waypoints, assigning the next waypoint to be the goal, then using a Laplacian path planner to construct a desired finely detailed path towards the goal, through the domain that avoids boundaries and objects in the domain. Assigning a potential value of v(r)=0 for r on boundaries and obstacles. Assigning a potential value of v(r)=−1 for r on a goal region, wherein the goal is a point on a planned path. Obtaining a numerical solution to the desired path with a Laplace's equation by gridding up the domain with a multi-sized cell grid, wherein the cells near an object are denser then the cells away from the objects. Iteratively setting a potential at each interior point equal to the average of its nearest neighbors and following the numerical solution provided by the Laplace's equation to the goal region.

A clustering method for high-dimensionality data includes identifying a set of nearest neighbors of a point in a multidimensional space and determining the centroid of the set of nearest neighbors, where the centroid is a member of the set of nearest neighbors. The method is then repeated using the neighbors identified around the computed centroid. In one embodiment, the method may terminate when the computed centroid becomes stationary over successive iterations. The resulting centroid may be returned as a mode of the data set. Points of the data set having common modes may be assigned to the same cluster.

The invention is a computer implemented technique for id entifying anomalous data in a data set. The method uses cascaded k-Means clustering and the ID3 decision tree learning methods to characterize a training data set having data points with known characterization. The k-Means clustering method first partitions the training instances into k clusters using Euclidean distance similarity. On each training cluster, representing a density region of normal or anomaly instances, the invention builds an ID3 decision tree. The decision tree on each cluster refines the decision boundaries by learning the sub-groups within the cluster. A test data point is then subjected to the clustering and decision trees constructed form the training instances. To obtain a final decision on classification, the decisions of the k-Means and ID3 methods are combined using rules: (1) the Nearest-neighbor rule, and (2) the Nearest-consensus rule.

The present invention relates generally to a system and method for categorization of strings of words. More specifically, the present invention relates to a system and method for normalizing a string of words for use in a system for categorization of words in a predetermined categorization scheme. A method for adaptive categorization of words in a predetermined categorization scheme may include receiving a string of text, tagging the string of text, and normalizing the string of text. Normalization may be performed with a three-stage algorithm including a literal match processing stage, an approximation match processing stage, and a nearest neighbor match processing stage. The normalized string of text can be compared to a number of sequences of text in the predetermined categorization scheme.

An integrated and unified method of statistical-like analysis, scenario forecasting, risking sharing, and risk trading is presented. Variates explanatory of response variates are identified in terms of the "value of the knowing." Such a value can be direct economic value. Probabilistic scenarios are generated by multi-dimensionally weighting a dataset. Weights are specified using Exogenous-Forecasted Distributions (EFDs). Weighting is done by a highly improved Iterative Proportional Fitting Procedure (IPFP) that exponentially reduces computer storage and calculations requirements. A probabilistic nearest neighbor procedure is provided to yield fine-grain pinpoint scenarios. A method to evaluate forecasters is presented; this method addresses game-theory issues. All of this leads to the final component: a new method of sharing and trading risk, which both directly integrates with the above and yields contingent risk-contracts that better serve all parties.

An RFID system and method for communicating between a host computer, one or more interrogators connected to the host computer, and a large body of transponders distributed within an area covered by the interrogators. Each transponder originally has a common identification code, and upon initialization by the host computer internally generates a unique identification code based upon an internally generated random number. The host, through the interrogators, reads each of the identification codes associated with each transponder by iteratively transmitting a read identification code command along with a controlled variable. Each transponder compares the received controlled variable to an internally generated random number, and selectively transmits its identification code based upon the outcome of this comparison. After the completion of each read identification code iteration, the host adjusts the controlled variable based upon the responses received in the previous iteration. Preferably, communications between the interrogators and the transponders are DSSS signals in TDMA format, and the transponders use the random number generator to assign a time slot for transmission of their response. Each interrogator includes an antenna system utilizing a switch matrix to connect multiple antennas having different polarizations, which ensures that all transponders within the range of the interrogator receive the signals from the interrogator. In a further aspect, the interrogators are arranged in groups, each group in nearest neighbor format, to reduce the time for reading the transponders and the emissions generated when more than one interrogator is active at the same time.

The invention provides a binocular camera-based panoramic image splicing method. According to the method, a binocular camera is arranged at a certain point of view in the space, the binocular camera completes photographing for once and obtains two fisheye images; a traditional algorithm is improved according to the defect of insufficient distortion correction capacity of a latitude-longitude correction method in a horizontal direction; corrected images are projected into the same coordinate system through using a spherical surface orthographic projection method, so that the fast correction of the fisheye images can be realized; feature points in an overlapping area of the two projected images are extracted based on an SIFT feature point detection method; the search strategy of a K-D tree is adopted to search Euclidean nearest neighbor distances of the feature points, so that feature point matching can be performed; an RANSAC (random sample consensus) algorithm is used to perform de-noising on the feature points and eliminate mismatching points, so that image splicing can be completed; and a linear fusion method is adopted to fuse spliced images, and therefore, color and scene change bluntness in an image transition area can be avoided.

An entangled links mechanism to establish and maintain bipartite temporal intimacy between pairs of computers, using an idempotent, reversible token method which presents no observable external “change” until a communication of information needs to occur between the computers, and which maintains the potential for “bounded (or unbounded) reversibility” in case the intended information dispatched by a source computational entity is not captured or properly accepted by a destination computational entity. The mechanism enables distributed computers in a network to remain continuously aware of each other's presence; to communicate on a logically nearest neighbor basis in a secure and reliable manner in which packets passed over these links do not conflict with normal traffic or cause the available resources of the link to be exceeded; and that atomicity, consistency, isolation, and “reversible durability” may be maintained for transactions when perturbations occur.

The invention relates to a method and a device for performing personalized recommendation on users. The method mainly comprises: calculating the similarity of users through label data of the users, calculating the similarity of products through label data of the products, and obtaining a neighboring set of each user and each product according to the similarity information; calculating characteristic vectors of the users and the products through a nearest neighbor matrix decomposition algorithm on the basis of the neighbor set information of the users and the products; and forecasting grading data of products not judged by the users according to the characteristic vectors of the users and the products, and performing personalized product recommendation according to the grading data. Labeling information and the neighbor set information can be guided into a matrix decomposing model by means of an embodiment of the method and the device, so that similar users/products can have the similar characteristic vectors, and the device for performing personalized recommendation on users can have the advantages of the matrix decomposing model and can capture the strong ties among neighbors in an overall situation.

An InfiniBand switch fabric is described. This InfiniBand switch fabric includes a hierarchy of switches having physical ports. These physical ports have associated local routing tables for nearest-neighbor connections to the physical ports. Moreover, relationships between the nearest-neighbor connections to each of the physical ports specified in the local routing tables are based in part on network-layer global identifiers that are assigned based on the hierarchy. Furthermore, local routing tables in the switches facilitate paths or routes for packets through the InfiniBand switch fabric without using a global routing table in the switches. In addition, the InfiniBand switch fabric includes a fabric manager, coupled to the switches, having the global routing table to assign the network-layer global identifiers.

The invention discloses a human face recognition method which is based on Gabor information fusion of visible light and near-infrared light. The method comprises: human face images under a visible light source and a near-infrared light source are respectively collected, Gabor features of the two images are respectively extracted to be fused at a feature layer; an AdaBoost algorithm is adopted to carry out the feature selection on the feature after the fusion, and the nearest neighbor classification is adopted to carry out the calculation and the classification on the similarity thereof. The human face recognition method has very high accuracy rate and excellent robustness of the impacts of light illumination on the human face recognition; in addition, compared with other methods, the human face recognition method has the advantages of small number of the used features, rapid classification speed, etc.

The invention discloses a method for splicing a video in real time based on multiple cameras, which comprises the following steps of: acquiring synchronous multi-path video data; preprocessing frame images at the same moment; converting a color image into a grayscale image; enhancing the image, and expanding the dynamic range of grayscale of the image by a histogram equalization method; extracting the characteristic points of corresponding frames by using a speeded up robust features (SURF) algorithm; solving matched characteristic point pairs among corresponding frame images of the video by using a nearest neighbor matching method and a random sample consensus matching algorithm; solving an optimal homography matrix of initial k frames of the video; determining splicing overlapping regions according to the matched characteristic point pairs; taking a homography matrix corresponding to a frame with highest overlapping region similarity as the optimal homography matrix, and splicing subsequent video frame scenes; and outputting the spliced video. The method can reduce the calculated amount of splicing the video frame single-frame image, improves the splicing speed of traffic monitoring videos and achieves real-time processing effect.

Illness signatures are mathematically characterized by entrainment relationships among multiple time series representations of physiological processes. Such characteristics include time and phase lags, window lengths for optimum detection, which time series are most entrained with each other, the degree of entrainment relative to the rest of the large database, and the concordance or discordance of the time-varying changes. These optimum disease-specific characteristics can be determined, for example, from large, clinically well-annotated databases of time series representations of physiological processes during health and illness. These characteristics of the entrainment relationships among multiple time series representations of physiological processes are used to make mathematical and statistical predictive models using multivariable techniques such as, but not limited to, logistic regression, nearest-neighbor techniques, neural and Bayesian networks, principal and other component analysis, and others. These models are quantitative expressions that transform measured characteristics to the probability of an illness, or p(illness).

The invention discloses an Android device wireless fidelity (WiFi) indoor locating method based on a position fingerprint identification algorithm. The Android device WiFi indoor locating method based on the position fingerprint identification algorithm aims at solving the problems that when a terminal receiver works in a city with dense building clusters or indoors, signal strength is greatly reduced due to influence by buildings, thus locating accuracy is low and location even can not be completed. Based on the foundation of the traditional position fingerprint identification algorithm, a k-nearest neighbor (KNN) matching algorithm and a coordinate computing method based on weight are adopted so that the position of a point to be measured is obtained, errors brought by signal fluctuation is effectively reduced, the position of the terminal receiver can be accurately located, and reaction is fast. Compared with a tradition indoor locating method, the Android device WiFi indoor locating method based on the position fingerprint identification algorithm can accurately locate the position of a requester in the condition of a complex environment, and is fast in reaction, efficient, accurate and especially suitable for indoor position location of an Android device terminal.

The invention provides parametric modules called Self Implementing Modules (SIMs) for use in programmable logic devices such as FPGAs. The invention further provides tools and methods for generating and using SIMs. SIMs implement themselves at the time the design is elaborated, targeting a specified FPGA according to specified parameters. In one embodiment, a SIM automatically places and interconnects child SIMs in a mesh pattern. The mesh is a 2-dimensional object corresponding to an array of CLBs on an FPGA. In essence, this embodiment allows a SIM to reserve routing resources on a target device (e.g., an FPGA), and allocate these resources to its child SIMs. Using a defined protocol, each child SIM can request and reserve routing resources, as well as placement resources (such as flip-flops and function generators in the CLBs) through the parent SIM. The routing resources are not necessarily limited to local or nearest neighbor routing.

A method for locating neighbor objects of a query object in a database. A query tile set is defined including a plurality of tiles that define an initial query radius that the query object is grouped within. Neighbor objects are located within the initial query radius. A number of neighbor objects within the initial query radius is determined. The query radius is expanded beyond the initial query radius by defining an expanded query tile set beyond the query tile set if the number of neighbor objects is less than a target number. Neighbor objects are located within the expanded query radius. It is determined whether a number of neighbor objects within the expanded radius corresponds to a target number. The query radius is expanded and neighbor objects located within the expanded radius until the number of neighbor objects equals or exceeds the target number.

A computer vision/image processing method generates a depth map useful in producing a foreground depth mask for 2D/3D image editing. The method uses image data from a plurality of scenes. Feature points on each of the vertical scan lines in each of the scene images are used to search for corresponding feature points on the corresponding vertical lines in other images. The corresponding feature-point search is performed by using a bipartite match network with a feature-point-ordering constraint and a disparity-limit constraint, and produces an individual feature-point depth map for each input image. A sparse feature-point depth map of the scene is obtained after applying a consistency test to all the individual depth maps. A complete feature-point depth map is produced by applying a color property assisted depth propagation process to the sparse feature-point depth map. Foreground and background separation is then conducted in the depth domain by using the order statistics of the depth data extracted the feature-point depth map. A foreground feature-point depth map is obtained from the separation operation. The final foreground depth mask is generated by applying a color aided eight-nearest-neighbor LMS interpolation process to the foreground feature-point depth map.

A method is provided for deriving optimized operating parameter settings for industrial furnaces of different designs as commonly used in power generation that will achieve robust and desirable operations (for example, low NOx and low CO emissions while maintaining specific furnace exit gas temperatures). The method includes the application of recursive partitioning algorithms to historical process data to identify critical combinations of ranges of operational parameter (combinations of settings) that will result in robust (low-variability) desirable (optimized) boiler performance, based on empirical evidence in the historical data. The method may include the application of various algorithms for recursive partitioning of data, as well as the consecutive application of recursive partitioning methods to prediction residuals of previous models (a methodology also known as boosting), as well as the application of other prediction algorithms that rely on the partitioning of data (support vector machines, naive Bayes classifiers, k-nearest neighbor methods).

The invention discloses a data-driven and task-driven image classification method. The data-driven and task-driven classification method comprises the steps that a convolutional neural network structure is designed according to the scale of data sets and image content; a convolutional neural network model is trained through the given classified data sets; feature expression is extracted from training set images through a trained convolution neural network; images to be tested are input into the trained convolutional neural network and are classified. The data-driven and task-driven image classification method is based on nonlinear convolution feature learning, and the model can be adapted to the data sets through a date driving mode, so that the specific data set can be better described; errors of K-nearest neighbors can be directly optimized through a task-driving mode, and therefore a better performance can be obtained with respect to a K-nearest neighbor task; efficient training can be conducted through a GPU in the training stage, and efficient K-nearest neighbor image classification can be achieved just through a CPU in the testing stage; in this way, the data-driven and task-driven image classification method is quite suitable for a large-scale image classification task, a retrieval task and the like.