21 results about "Sample number" patented technology

The invention discloses a report checking method capable of automatically feeding back a checking result based on a snapmake language. The method specifically comprises the following steps: step 1, file preparation: preparing a question settlement report and an analysis confirmation sheet excel table; 2, performing report basic information checking, quality control data checking, intra-group sample difference checking, report sample number and display sequence checking, report integrity checking and / or report picture checking in parallel and / or in series on the basis of the question settlement report obtained in the step 1 and an analysis confirmation single excel table; and step 3, generating a webpage version checking report according to one or more pieces of checking log information obtained in the step 2, and automatically sending feedback through a mail. The invention further discloses a system and application for implementing the report checking method, and the system and the application have wide application value.

The invention discloses a model increment training method and device based on dynamic storage, a storage medium and a program product, and relates to the technical field of artificial intelligence, and the model increment training method based on dynamic storage comprises the steps: obtaining at least one piece of original sample data, and determining an information amount score corresponding to the original sample data; determining the original sample data with the information content scores higher than a preset score threshold as training sample data, storing the training sample data in a preset memory, and accumulating the information content scores of the training sample data in the preset memory to obtain a corresponding accumulated information content score; and performing incremental training on the to-be-trained model by utilizing the training sample data in the preset memory to obtain a target model in response to the condition that the accumulated information amount score or the sample number corresponding to the training sample data in the preset memory meets a preset triggering condition. The technical problem that the training effect of a traditional field small model is affected due to data distribution offset of training data is solved.

The present disclosure relates to a text processing method and device, equipment and storage medium, and belongs to the technical field of computer, which can improve the accuracy of text processing. The text processing method comprises: obtaining a to-be-modified text; determining a target text according to the to-be-modified text, a preset number of words and a pre-trained text modification model; the number of words of the target text is the preset number of words; the text modification model is obtained according to a sample text pair and a sample number; the sample text pair comprises a to-be-modified sample text and a modified sample text; and the sample number is used to represent the remaining number of words after the byte corresponding to the sample number in the modified sample text.

The invention relates to the technical field of data processing, in particular to a processing method for task classification and subclass sample enhancement of a corner plastic data set, which comprises the following steps of: S1, acquiring a label set of the corner plastic data set, judging whether each label in the label set conforms to a classification task or a regression task, and if yes, judging whether each label conforms to the classification task or the regression task; taking samples corresponding to the same label in the corner plastic data set as a label category; s2, a learning model is constructed for the labels conforming to the classification task, the learning model adjusts the sample weight of a sample corresponding to each label category through a dynamic category weight method, and then the sample number of the angle plastic data set is adjusted through a self-adaptive mixed sampling method to form an adjusted data set; s3, adjusting the attention degree of the samples in the adjusted data set; and S4, performing prediction fusion on the adjusted data set processed in the step S3 through a subclass gating fusion method, and further executing a classification task on each sample in the adjusted data set. According to the invention, the sample processing accuracy is improved.

An apparatus, a method, and a computer program product for sidelink positioning are provided. An example method may include receiving a set of sidelink control information (SCI), where the set of SCIs reserves a set of resources for transmission of a set of sidelink positioning reference signals (SL-PRSs) in at least one same slot, where the set of resources is periodic. The example method may also include measuring the set of SL-PRSs during the measurement period based on the set of SCIs satisfying at least one of the set of conditions.

This invention provides a method, apparatus, electronic device, and storage medium for identifying credit card fraud. The method includes: determining target numbers corresponding to behavioral characteristics from a set of numbers to be identified based on indicator data of a sample number set; and determining fraudulent numbers based on the target numbers. The behavioral characteristics include joint fraudulent behavior characteristics, and the target numbers corresponding to these joint fraudulent behavior characteristics are numbers corresponding to a joint fraudulent group. The joint fraudulent group is obtained by comparing the number trajectory similarity of the set of numbers to be identified with a similarity threshold corresponding to the number trajectory similarity of the set of numbers to be identified. The similarity threshold is obtained through iterative training based on the number trajectory similarity of the sample number set, the indicator data, and a loss function of a greedy algorithm. This invention improves the accuracy and efficiency of credit card fraud identification.

The invention relates to the technical field of fishery resource sample management, in particular to an intelligent network library retrieval and management method and system for fishery resource samples. Multidimensional sample information is collected, and a label coding layer is constructed to paste a unique retrieval label for each sample; a category search tree is constructed by combining a cabinet body structure of a sample storage cabinet and a classification system, three-level codes of cabinet number-column number-box number are formed, and standard classified storage of fishery resource samples is achieved. Sample information is input into an Excel table data network library based on inverted indexes, and multi-condition rapid retrieval of places, time, sample numbers and the like is achieved. Before warehousing, three samples are randomly extracted from each box to verify the consistency of the labels and the database, and no less than 5% of samples are regularly backed up and randomly checked, so that long-term accuracy of the data is ensured. According to the method, the fishery resource samples can be classified, stored and taken for rapid positioning and retrieval, the cost is low, the operability is high, and the sample positioning, retrieval and management efficiency is remarkably improved.

The invention discloses a contact window detection method and device based on dynamic density clustering decision. The method comprises the following steps: collecting multi-source physical signals, and performing weighted summation on the multi-source physical signals based on a compensation coefficient to obtain a synthetic Tsyn signal; dividing the continuous synthetic Tsyn signal into a plurality of equal-width intervals and generating a histogram; extracting a maximum count value from the histogram and multiplying the maximum count value by a signal-to-noise ratio correlation coefficient k to serve as a main density threshold value; comparing the main density threshold value with a 0.05 quantile value of the synthesized Tsyn signal, and taking the larger value as a dynamic density threshold value; extracting a section which meets a dynamic density threshold and has a duration exceeding M in the continuous synthetic Tsyn signal as a feature sequence; judging the sample number of the signal amplitude in the feature sequence, and if the sample number is not greater than 5, processing and identifying the contact section by adopting a statistical mode; and if the number of the samples is greater than 5, processing and identifying the contact segment by adopting a clustering mode. According to the invention, real-time accurate identification of the contact window can be realized.

The invention relates to a biological detection data analysis method, which belongs to the technical field of data analysis, and comprises the following steps: obtaining inspection data, generating a unique sample number, calling a standard template and matching a standard detection item, generating a sample preparation guide, receiving a triggered preparation confirmation instruction, obtaining a sample preparation completion instruction, and triggering a sample detection instruction. The method comprises the following steps: acquiring original data of equipment, judging according to a judgment standard to obtain a detection result, calling a multi-dimensional data analysis engine to perform statistical analysis on the detection result based on the detection result, and calculating an epidemic disease positive rate and comparing the epidemic disease positive rate with a preset positive rate reference range to obtain an analysis result, so that intelligent upgrading in the field of biological detection is realized; the traceability and the processing normalization of sample information are ensured, the problems of large personal error and data analysis lagging in the traditional detection process are effectively solved, and the comprehensive purpose of improving the detection accuracy, the detection efficiency and the risk early warning capability is achieved.

The invention relates to the technical field of medical data processing and intelligent analysis, and discloses a breathing alarm monitoring system and a processing method thereof. The method comprises the following steps: setting a sampling period and monitoring duration, calculating a sample number and forming an original sequence, calculating a mean value and a standard deviation, performing complete sequence standardization, and setting a numerical value stable constant; if the sample number is zero or one, an empty result is output. Then setting a segment number range and a minimum segment length, generating candidate segments based on an equidistant reference and recursive constraint, and calculating a mean value and a variance of each segment; constructing a coding length containing scale penalty and fitting error, and superposing the segment number and the combination item to obtain an overall coding length; and selecting an optimal one among the candidates according to a parallelization rule. And standardizing the mean difference of the adjacent segments, setting a threshold value for alarm judgment when the number of the segments is enough, mapping segment points into time, and outputting a structured alarm.

1. Name of the product in this design: Operation area of ​​a graphical user interface for inputting sample information in an electronic device. 2. Intended use of this design: for use in electronic devices, with the operating area for inputting information required for measuring samples such as bacteria. 3. The key design features of this product are the solid line drawings in the graphical user interface. 4. The image or photograph that best illustrates the design's key features: the front view. 5. Since the key design feature of this design is the graphical user interface, the rear view, left view, right view, top view, and bottom view are omitted. 6. Uses of the graphical user interface: to input information needed to measure samples such as bacteria, or to display images of petri dishes used as containers for culturing bacteria. 7. Graphical User Interface: Users can select the desired measurement type by choosing options such as "Examine" or "Prevent Measurement," and input various information into input fields such as "Sample Number," "Measurement Object," and "Cultural Plate" for the sample being measured. When the "Start Measurement" button at the bottom of the information input area is clicked, the measurement of the sample begins, and an image of the culture dish, which serves as the container for culturing bacteria, is displayed on the image display area on the left side of the interface.

The invention discloses a big data cleaning method and system for an intelligent electric energy meter. The method comprises the following steps: loading original test data of the intelligent electric energy meter; performing three-layer grouping and encoding on the original test data according to a test scheme, manufacturers and sample numbers, and encoding applied electric stress; decomposing the coded input time data into different time dimensions and carrying out normalized coding; carrying out abnormal value detection and elimination on the input time data after normalization coding by adopting a weighted isolation forest algorithm based on feature importance; filling missing values by adopting an improved K-nearest neighbor decision tree algorithm; and storing the processed data according to a three-layer grouping mode of a test scheme, a manufacturer and a sample number. The invention aims to realize automatic and standardized data cleaning and coding, provide ideal input for a deep learning model, and improve the accuracy of operation state evaluation and degradation prediction of the intelligent electric energy meter.

The invention relates to the technical field of network security, and discloses an outbound C2 detection method based on a flow level risk score and destination end entity aggregation, which comprises the following steps of: firstly, acquiring outbound packet-by-packet records containing timestamps, quintuples and packet lengths; maintaining a session state according to a quintuple, and generating a session sample when a packet number triggering condition and a time window triggering condition are met; semantically extracting an arrival interval and a packet length sequence according to an internal network side direction, and executing negative value clamping, threshold cutting and log1p compression on the arrival interval to form session features; inputting a risk scoring model to obtain a flow-level risk score; aggregating the multi-stream scores according to target end entity keys to obtain entity-level risk scores; and obtaining a calibration set, triggering alarm according to a fixed false alarm rate constraint when the number of samples is not less than a threshold N, triggering alarm according to an alarm budget constraint when the number of samples is less than the threshold N, and outputting an evidence chain. According to the invention, the alarm quantity is controllable, the threshold value is adjustable and the processing cost is reduced under the non-decryption condition.

This image is a test selection image for setting the test details for a sample in an automatic analyzer that tests samples (e.g., blood or urine). A list of sample numbers is displayed on the left side of the screen, corresponding to their positions in the sample rack within the automatic analyzer. Detailed information, such as the type of sample selected from the list of sample numbers on the left, is displayed in the upper center section of the screen. The middle center section of the screen displays the test items that can be measured for the selected sample, and the test details are set by selecting a test item. An area for displaying operation buttons, such as a save button, is provided in the lower center section of the screen.

The invention relates to an analytical machine arrangement (100). A method comprises obtaining (A10) N time domain measurement samples at a time T (n), where n traverses the number of samples N, and the distance between adjacent samples in the time domain is T. The method further comprises determining (A12) updated, normalized time-domain measurement samples having a timing Tadapt (n) for each sample based on the associated T (n) and T adjustment, further based on an actual frequency factual (n) of the machine arrangement (100), and further based on an undesired behavior and / or error-related frequency F. The method further comprises determining (A12) the updated, normalized time-domain measurement samples having the timing Tadapt (n) for each sample based on the associated T (n) and T adjustment, further based on the actual frequency factual (n) of the machine arrangement (100). The method further comprises determining (A14) M reconstructed, normalized time domain measurement samples having a constant sampling period T2 at a time T (m), where m traverses the number M of samples. Furthermore, the method comprises performing (A16) a frequency analysis on the reconstructed time-domain measurement samples.

The invention relates to an electric heating coil accelerated life test method and system, and relates to the technical field of electrical component testing, and the method comprises the steps: obtaining the performance parameters of a plurality of initial samples in response to a preset test signal; initial samples are obtained and numbered randomly to obtain initial sample numbers, and the initial samples are screened according to the performance parameters and a preset screening rule to obtain performance sample numbers; obtaining a sample stress group based on the performance parameters and the screening rule, wherein the sample stress group is voltage-temperature dual-stress coupling; controlling the electric heating coil corresponding to the performance sample number to perform a preset orthogonal test according to the sample stress group to obtain a test result; inputting the test result into a preset standard life model for training to obtain a correction parameter; and correcting the standard life model by using the correction parameter to obtain an actual life model, and outputting the actual life model. The method has the advantages that test parameters are unified, steps are standardized, and test results are accurate and reliable.

The application discloses a kind of based on constant correlation and low sample number Doppler signal fast detection method, comprising: the signal of receiving end collection is carried out Hilbert transform, obtains analytic form received signal, and with the same down-sampling multiple of the constant correlation detection template constructed in equal interval extraction, obtain low sample detection signal sequence;With the low sample template signal in constant correlation detection template, the normalized cross-correlation calculation of down-sampled received signal is carried out, extracts the maximum peak value of correlation function and its corresponding Doppler compression ratio;According to the Doppler compression ratio corresponding to maximum correlation peak, the Doppler factor and Doppler frequency offset parameter of signal are calculated, and the detection result is output, realize the fast estimation of underwater acoustic signal Doppler signal.The down-sampling process of the application will not change the amplitude and position of cross-correlation main peak, so as to significantly reduce the sample number and the calculation complexity, while maintaining the detection precision and Doppler signal detection performance.

This application relates to the field of equipment and item recognition technology, and discloses an intelligent item recognition method and apparatus based on long-tail dynamic weight adjustment. The method includes: acquiring an image sample of an item to be recognized; inputting the image sample into a trained item recognition model; the item recognition model is a student model trained based on differential distillation constraints applied to sample frequencies of different item categories; and outputting the category recognition result of the item to be classified. Since the item recognition model is a student model trained based on differential distillation constraints applied to sample frequencies of different item categories, distillation constraints of different intensities can be applied to item categories with different sample numbers, thereby enhancing the student model's learning ability for item categories with fewer samples, mitigating the adverse effects of uneven distribution of category samples on recognition performance, and improving the accuracy of item recognition.

The invention discloses a table data synthesis method, system and device and a storage medium, and the method comprises the steps: carrying out the deduplication of all semantic dimension label values of a first table training sample set, combining each semantic dimension of each first table training sample and each semantic dimension label value of each semantic dimension, and obtaining a table data synthesis result; deduplication is carried out on first combined label values in the first combined label value set, the occurrence frequency of each second combined label value in the second combined label value set in the first combined label value set is determined, and on the basis of the total frequency of each second combined label value, the first probability value and the preset training sample number, the number of training samples in the first combined label value set is determined; the first table training sample set is screened, the first table data synthesis model is trained based on the second table training sample set, the trained table data synthesis model is obtained, and the table data synthesis efficiency and accuracy are improved.

The application discloses a method and system for determining magnification and minimum sample number in micro-component content analysis, and relates to the technical field of microscopic image quantitative analysis. The method comprises the following steps: constructing a standardization simulation model, excluding the interference of natural rock heterogeneity, and simulating the actual microscopic observation process by adopting a double-mode sampling method; using a double statistical stability determination method to objectively determine the minimum observation sample size under different magnifications, establishing a quantitative correlation between the magnification and the minimum sample number, and determining the optimal observation magnification by taking into account the analysis accuracy and efficiency. The application realizes the standardization of sample selection and the controllability of statistical error in micro-component quantitative analysis, effectively eliminates subjective bias caused by artificial experience, significantly improves the accuracy, stability and repeatability of the component content analysis results, and can be widely applied to rock and mineral identification, reservoir evaluation and microscopic image panoramic stitching scenes, thereby providing a standard and reliable technical support for rock multi-scale micro-component quantitative analysis.

Embodiments of the present application provide a sample library updating method and device of a self-learning coefficient calculation model, and relate to the technical field of rolling force calculation model control. The method comprises: determining a minimum effective capacity; creating a plurality of sample libraries; judging whether the sample libraries meet the requirements based on the minimum effective capacity; calling the sample libraries that meet the requirements, and simultaneously updating other sample libraries. The present application periodically updates the sample data of the sample libraries, guarantees the "freshness" of the sample number, improves the application accuracy of the self-learning coefficient when applied to variable-gauge production, avoids the influence of the original accumulated self-learning coefficient on the current production due to changes in equipment parameters, and further makes the steel rolling production more stable and smooth.