45 results about "Sparse data sets" patented technology

An implementation of NMF functionality integrated into a relational database management system provides the capability to apply NMF to relational datasets and to sparse datasets. A database management system comprises a multi-dimensional data table operable to store data and a processing unit operable to perform non-negative matrix factorization on data stored in the multi-dimensional data table and to generate a plurality of data tables, each data table being smaller than the multi-dimensional data table and having reduced dimensionality relative to the multi-dimensional data table. The multi-dimensional data table may be a relational data table.

Disclosed herein is a method of analyzing large scale survey results comprising obtaining a sparse data set representing a subset of an original data set comprising a plurality of individuals' responses to a plurality of questions, wherein the sparse data set comprises less than ninety percent of the responses in the original data set; analyzing the sparse data set using a general diagnostic model; and obtaining estimated person parameters using the general diagnostic model.

The invention relates to compressing of sparse data sets contains sequences of data values and position information therefor. The position information may be in the form of position indices defining active positions of the data values in a sparse vector of length N. The position information is encoded into the data values by adjusting one or more of the data values within a pre-defined tolerance range, so that a pre-defined mapping function of the data values and their positions is close to a target value. In one embodiment, the mapping function is defined using a sub-set of N filler values which elements are used to fill empty positions in the input sparse data vector. At the decoder, the correct data positions are identified by searching though possible sub-sets of filler values.

A method for training a classifier for selecting features in sparse data sets with high feature dimensionality includes providing a set of data items x and labels y, minimizing a functional of the data items x and associated labels yL⁡(w,b,a,c,γ1,γ2):=1N⁢∑i=1N⁢ai+λ1⁢c1+λ22⁢w22+γ1T⁡(e-Y⁡(Xw+be)-a)+γ2T⁡(w-c)+μ12⁢e-Y⁡(Xw+be)-a22+μ22⁢w-c22to solve for hyperplane w and offset b of a classifier by successively iteratively approximating w and b, auxiliary variables a and c, and multiplier vectors γ1 and γ2, wherein λ1, λ2, μ1, and μ2 are predetermined constants, e is a unit vector, and X and Y are respective matrix representations of the data items x and labels y; providing non-zero elements of the hyperplane vector w and corresponding components of X and Y as arguments to an interior point method solver to solve for hyperplane vector w and offset b, wherein w and b define a classifier than can associate each data item x with the correct label y.

A system and method is provided for acquiring a medical image of a portion of a vascular structure of a subject using a magnetic resonance imaging (MRI) system. At least one radio frequency (RF) saturation pulse is applied to a selected region of a subject that is free of exogenous contrast agents using the MRI system to saturate spins of all tissues within the selected region. A delay time is selected to allow an inflow of unsaturated vascular spins that are free of influence from exogenous contrast agent into the selected region through vascular structures within the selected region. A sparse dataset is formed from a series of spatially-encoded views from the selected region using an undersampled radial k-space trajectory in which the center of k-space is sampled for each view in the series of spatially-encoded views. An image of the vascular structures within the selected region is reconstructed from the sparse dataset.

The invention discloses a wireless sensor network (WSN) crop growth environment information sparse acquisition and transmission method based on NB-IoT and FPGA and applied to smart agriculture. The method mainly comprises the steps that crop growth environment information is collected in a sparse sampling mode through Zigbee wireless sensor nodes; the collected sparse data information is sent to an FPGA core control end, and a matrix filling algorithm is executed in the FPGA to fill the sparse data set matrix; data recovered to be complete is sent to the cloud platform end in the NB-IoT mannerand is displayed at the cloud end, so that the functions of collecting, transmitting and monitoring the crop growth environment information are finally realized. According to the invention, a sparsesampling theory is applied to the wireless sensor network, and a matrix filling algorithm is realized by hardware, so that lower power consumption of sensor nodes and higher data recovery speed can beensured.

The invention relates to a recommendation method and system of efficient data analysis. Specifically, the synchronous real-time recommendation of a scene perception recommendation algorithm is realized through a grading module, a scene optimization module and a big data operation module. A scene model is introduced to correct a sparse linear method, and a recommendation engine based on memory calculation is imported, so that recommendation efficiency is greatly improved. Meanwhile, the problems of the algorithm optimization of a sparse data set and recommendation result abduction are solved. The system can be applied to the real-time recommendation of supermarket consumption. Compared with existing off-line recommendation, the system is characterized in that a user can timely receive relevant recommendations after selecting or purchasing corresponding products, and therefore, a conversion rate brought by the recommendation system is greatly improved. A scene model is imported to cause the user to obtain accurate recommendations under the certain scene under different scenes, and user experience is improved.

The invention discloses a privacy protection association rule mining method based on a sparse data set, and mainly solves the problems that an existing mining method needs to introduce additional calculation overhead and privacy leakage. The scheme comprises the following steps: 1) initializing a system; 2) enabling the data owner to encrypt the uploaded data; 3) enabling a data miner to encrypt, upload and query; 4) performing double-cloud calculation on an inner product ciphertext set of query and transaction, and calculating a support degree ciphertext of query; 5) comparing the support degree with a support degree threshold value by the double-cloud security; (6) enabling the data miner to decrypt the mining result, judge whether the mining result is a frequent item set or not, solve a non-empty proper subset of the frequent item set, and querying, encrypting and uploading the non-empty proper subset and the query; 7) comparing the confidence coefficient with a confidence coefficient threshold value by the double-cloud security; and 8) enabling the data miner to decrypt the mining result and judge whether the mining result belongs to the strong association rule. According to the invention, the risk of privacy disclosure is reduced, and higher privacy protection requirements can be met; and meanwhile, the mining calculation efficiency is effectively improved.

A scalable configurable universal full spectrum cyber process that utilizes measure points from sensor observation-derived representations or analytically rich sparse data sets for making selected cyber determinations regarding or utilizing sensor observations or sensor observation subjects. Utilizing necessary cyber resources and predetermined criteria for making selected cyber determinations regarding or utilizing the sensor observation or at least one sensor observation subject, the disclosed cyber process employs the use of measure points for accurately or reliably locating selected analytically rich indicators from sensor observation-derived representations, wherein appropriate informational representations or measurements regarding or utilizing the selected analytically rich indicators, the measure points or the sensor observation are assigned and stored in analytically rich sparse data sets where they may be utilized by the cyber process for making, in real time or at any time thereafter, selected cyber determinations regarding or utilizing sensor observations or sensor observation subjects.