71 results about "Compositional data" patented technology

A high-density, distance-measuring laser system and an associated computer that processes the data collected by the laser system. The computer determines a data partition structure and stores that structure as a header file for the scan before data is collected. As the scan progresses, the computer collects data points until a predetermined threshold is met, at which point a block of data consisting of the data points up to the threshold is written to disk. The computer indexes each data block using all three coordinates of its constituent data points using, preferably, a flexible index, such as an R-tree. When a data block is completely filled, it is written to disk preferably with its index and, as a result, each data block is ready for access and manipulation virtually immediately after having been collected. Also, each data block can be independently manipulated and read from disk.

The embodiment of the application provides a data processing method and device. The method comprises following steps: obtaining preset configuration information including data element information and data processing information, and application layer description information; reading data from the data source location according to the data element information and processing the read data according to the data processing information; composing processed data into a data set; and processing the data set according to the application layer description information to obtain the applied data. The embodiment of the data processing method has following beneficial effects: when the data sources changes or business development needs to change processing logic of data, time is saved and labor and computing resources can be saved; and changes and updates of business can be rapidly responded to.

A method and apparatus for quantitative analysis of a material in which an electron beam is caused to impinge upon the material are described. The method comprises detecting low loss electrons (LLEs) received from a first region of the material due to interaction with the electron beam and generating corresponding LLE data. The method further comprises detecting x-rays received from a second region of the material due to interaction with the electron beam and generating corresponding x-ray data, wherein the first and second regions overlap, and analysing the LLE data together with the x-ray data so as to generate compositional data representative of the composition of the first region.

The LVARS is a fully instrumental, non-destructive spectroscopic device for the analysis and verification and authentication of the optical and electromagnetic properties (OEMP) of the inks, dyes, thin films, plastics, toners, paper, fixatives, paints, and printing agents used in documents and financial instruments. The instrument is quantitative in nature so as to correlate compositional data (elemental, isotopic, structure) with Raman optical spectra. The LVARS design consists of a computer-controlled spectrometer with a microscope-guided grid head containing the laser excitation source and detector and optics.

Systems and methods for mapping random data writes to sequential or semi-sequential storage. Data writes may be initially directed to an SSD or other relatively low latency and high IOPS data storage layer. The data may remain in the SSD storage layer until a group of data may be written, together, to a sequential or semi-sequential storage layer. A data organizer may group the data into data buckets based on tags and/or policies associated with the data. In this way, data subject to similar lifetime, priority, data protection, and/or other policies may be stored on a same segment or other portion of the sequential storage layer. Similarly, data having similar access patterns, authors, files, objects, or project affiliations, may be stored together on a same segment of the sequential storage layer.

The disclosure relates to a personalised interactive mail publishing method which can be used to consult personal data, trigger actions or acknowledge receipt in an instant and secure manner. The method includes: formation or extraction of the constituent data of a printable code; generation of complementary variables, such as for authentication; concatenation and compaction of the data from the code to be printed; formation of a file containing the data from the code to be printed and from the mail; conversion of the codes to enable optoelectronic capture; encapsulation of the code in a data flow; printing prior to the distribution of the mail or package; capture of the printed code by the recipient using, for example, a mobile telephone fitted with a camera; interpretation of the symbols forming the code using an embedded module; processing of the instructions contained in the code; connection to the server; encryption and comparison with the encryption performed by the embedded module; authentication; and access to the information indexed in the databases made available by the originator of the mail.

A video calendar system for providing a video calendar as a background image of an operating system (OS) with a graphical user interface (GUI) environment and applying a variety of variations to the video calendar according to a user's settings to provide novel pleasant images to the user. The video calendar system comprises an environment setup unit for setting up a video calendar output environment, a storage unit for storing moving image data, calendar constituent data, and environment configuration information set up by the environment setup unit, a video calendar creator for selecting moving image data and calendar constituent data from the storage unit on the basis of the environment configuration information set up by the environment setup unit and synthesizing the selected moving image data and calendar constituent data to create a video calendar to be reproduced on a background screen, and a display unit for displaying the video calendar created by the video calendar creator on the background screen, and a controller for controlling the entire system operation.

The invention discloses an efficient learning system based on a spiking neural network and a learning method thereof, and belongs to the technical field of artificial neural networks. The learning system is composed of a data preprocessing module, a network construction module, a learning module and an inference module. The data preprocessing module supports two types of static image data and dynamic neuromorphic data. The network construction module supports multiple types of spiking neural networks. The learning module is composed of feedforward network calculation, error back propagation and weight updating. At each moment, the random number generator generates a random value, and whether error back propagation and weight updating are carried out or not is controlled by a switch. The inference module loads the model file to carry out feedforward network calculation. The learning system is compatible with an automatic differential tool and is easy to implement, and various pulse neural network types are supported, and flexible configuration is easy. A switch structure is added to the learning module, a large number of calculation steps are ignored, calculation resources are saved, and the learning efficiency is improved.

A related information provision system may be used to identify constituent data for a selected portion of summary data on a summary data visualization such as a chart or graph. The portion of the summary data may be selected by a user, a system event, and/or another process. The constituent data may include values and/or metadata for one or more data sets summarized by the summary data, formulas and/or other information, which may indicate how the summary data are obtained from the constituent data. Information regarding the constituent data (“related information”) may be displayed for a user on a constituent visualization or the like. If the information is already displayed, the system may draw the user's attention to the information. If the information is displayed in a different view, the system may display a navigational element leading to it, or direct the user's attention to an existing navigational element.

The invention discloses a space-time evolution analysis method for a regional water conservation function. The method comprises the following steps: carrying out grid processing on collected data through the collected rainfall data, temperature data, DEM data, soil composition data and soil type data of a meteorological station, and carrying out normalization and ArcGIS software by adopting a maximum and minimum value method to obtain grid data of each factor; according to the grid data of each factor, on the basis of an NPP quantitative index method, calculating by combining ArcGIS software to obtain annual ecosystem water conservation service capability index grid data; analyzing according to the annual ecosystem water conservation service capability index grid data to obtain a time change trend of the water conservation service capability in the whole region space; and the time variation degree of the water conservation service capability in the regional space, the spatial aggregation characteristic of the water conservation function in the regional time and the time variation characteristic of the water conservation function in the different soil types of regions.

The invention relates to the technical field of engineering cost, in particular to a BIM-based whole-process engineering cost consultation system, which comprises a BIM model construction unit, a costmanagement unit and a user experience unit, and is characterized in that the BIM model construction unit is composed of an information data acquisition module, a BIM database module and a Revit modeling module; the BIM database module is in butt joint with the background database to transmit data, and the cost management unit comprises a project estimation module, a project rough calculation module, a construction drawing budget module, a completion settlement module, a real-time price collection module and a project progress control module. And the user experience unit comprises a real-timeprice change query module, a 3D model development module and a user feedback module, the real-time price change query module adopts a real-time information prompt mode, and the 3D model development module adopts a holographic projection technology. According to the invention, comprehensive consultation experience can be provided for customers, and the customers can know the real-time cost change situation in time.

The invention discloses an agricultural product price prediction method based on a seq2seq model and a CNN model, and the method comprises the steps: collecting historical price data of an agricultural product to form a data set M, carrying out the preprocessing of the data set M, converting the preprocessed data into a price trend image, and adding a label; Processing the data set M into multi-dimensional index data to train a seq2seq model; Using the price trend image with the label to train a CNN model; Processing the original historical price data of the to-be-predicted agricultural product into multi-dimensional index data, and inputting the multi-dimensional index data into the trained seq2seq model to obtain a quasi-prediction result Vs; And inputting the price trend image with thelabel into the trained CNN model to obtain a quasi-prediction result Vc, and correcting the prediction result Vs by using the quasi-prediction result Vc to obtain the final prediction price of the to-be-predicted agricultural product. Through double-model mixing, the method has high accuracy for agricultural product price prediction.

An information processing system includes: a data processing unit that executes verification processing for a content recorded in a disk and reproduces the disk-recorded content under a condition that the verification succeeds, wherein the data processing unit randomly selects hash units, which are objects of collation, from among a plurality of hash units formed with component data items of the content, reads the selected hash units sequentially from the disk, calculates hash values, and collates the calculated hash values with collation hash values; and the data processing unit executes reading sequence determination processing so as to determine a reading sequence in which the selected hash units are sorted according to recording positions in a disk, and reads the selected hash units according to the determined reading sequence.

The invention discloses a method of using SD information encryption to realize quick SD and AP connection in an uncorrelated WIFI environment. A third-party mobile terminal MT is also included. The method comprises steps: the SD virtualizes a pseudo AP; the MT actively scans nearby APs, the pseudo AP is identified through a near-distance wireless identification technology, MAC of the SD is acquired, and the MAC is sent to the AP; the AP uses the MAC to encrypt self device information to form a data flow, and the data are embedded in broadcasting frames to form a broadcasting packet to be sent; after the SD receives the broadcasting packet, the self MAC is used for decryption, the device information of the AP is extracted, correlation with the AP is thus carried out, and quick connection is completed. The third-party mobile terminal MT is delicately used as a medium to acquire the MAC information of the SD, the information thus serves as a key to encrypt the self device information of the AP, the AP can limit a receiving party while actively broadcasting the self device information, an error connection condition during an intelligent networking process can be avoided completely, the operation is accurate and reliable, and the security is high.

The invention relates to the technical field of power system load modeling, in particular to a method for selecting a typical load characteristic transformer substation based on multi-source data. Themethod comprises the steps of conducting load characteristic general survey on transformer substations under the same voltage level of a target power grid to obtain load characteristic data, whereinthe load characteristic data comprise load type data and industry composition data; performing type clustering analysis on the transformer substations according to the load type data, and performing load classification on the transformer substations to obtain a plurality of transformer substation groups with similar load characteristics; performing industry clustering analysis on the transformer substation group according to the industry composition data, and performing industry classification on the transformer substations; and selecting a typical transformer substation capable of representing load characteristics according to the load classification and the industry classification. According to the method, the aggregation theory method is applied to selection of the typical stations, thebasis for selecting the typical stations is provided according to the membership relationship, and a scientific basis is provided for modeling personnel in the process of selecting the typical stations.

The invention discloses an industrial process data clustering method for density peak clustering, which comprises: acquiring industrial process data to form a data set; combining the Euclidean distance between data in the data set with a time factor to obtain the distance between the data; obtaining a truncation distance according to the distance between the data and an adjustment parameter, so asto obtain the local density of each data, and calculating the minimum distances between each data and the data having local density larger than its local density; ordering the products of the local density of each data in the data set and the minimum distances, using the first H pieces of data with large products as clustering centers, and determining that the data, in the data with larger localdensity than the cluster center, closest to the clustering center and the clustering center belong to the same category; determining the class attribute of the data without class attributes in the data set according to a descending order of the local densities, so as to obtain the clustering result of industrial process data. The clustering centers of the invention are reasonable, and the number of clustering centers is automatically determined and less time complexity is achieved.

The invention discloses a cloud storage data system based on data self-destruction. The cloud storage data system includes a client, a metadata server and a plurality of storage nodes, wherein the client is used for encrypting, uploading, downloading and decrypting data and comprises a data encryption module, a Key splitting and reassembly module, a data decryption module and a communication module; the data encryption module is used for carrying out encryption processing on original data through a given key; the key segmentation and recombination module is used for segmenting the key in datauploading, generating a plurality of equal parts and randomly selecting part of or all of the equal parts from the equal parts to form a data object; The communication module is used for establishingconnection and data transmission between the client and each storage node; The data decryption module is used for decrypting the downloaded data through the regenerated key; The metadata server is used for generating metadata information of each divided key equal part on each corresponding storage node; And each storage node is used for creating a storage data object according to the established connection so as to store the encrypted data.

The invention relates to the technical field of machine learning, and provides a rapid clustering method based on density sub-graph estimation, computer equipment and a storage medium in order to overcome the defects that in the prior art, the mass center of a cluster cannot be determined, the calculation cost is high, and over-segmentation occurs in the clustering process. The rapid clustering method based on density sub-graph estimation comprises the following steps: acquiring samples, and preprocessing the samples to form a data set; performing density value estimation on each sample in thedata set, and constructing a density sub-graph set; finding out the highest density point of each density sub-graph from the density sub-graph set to serve as a representative point of the density sub-graph, and forming a candidate set by samples corresponding to the representative points; calculating an important value of each sample in the candidate set; sorting the candidate sets in a descending order according to the important values, and selecting the first K samples as the centroids of the K clusters; and classifying non-centroid samples in the candidate set, and outputting to obtain aclustering result.