173 results about "Reference dataset" patented technology

The invention belongs to the field of data quality analysis, and particularly discloses a data set quality analysis and evaluation method, which performs integrity evaluation on a current data set, starts data confidence analysis only after the number of data groups reaches the standard, ensures that confidence evaluation is established on the basis of data with a complete structure, avoids erroneous judgment caused by data missing, and improves the quality of the data set. Meanwhile, during data confidence analysis, dynamic comparison between current data and a local time sequence neighborhood is implemented by constructing an adjacent data set of each data group; meanwhile, a historical reference data set is combined, longitudinal comparison with a historical fluctuation range is carried out, and dual verification of local consistency and historical comparability is fused, so that the anomaly detection sensitivity and the evaluation accuracy are improved to a great extent; and finally, establishing a closed-loop feedback path with an acquisition end when availability judgment is carried out based on a data set integrity and confidence evaluation result, and realizing self-adaptive adjustment of an acquisition strategy according to a risk positioning result.

System and methods are disclosed for matching extracted text data based on one or more similarity scores. The method may include receiving one or more documents from a plurality of data sources, utilizing an optical character recognition algorithm for extracting text data from the one or more documents, comparing, utilizing a fuzzy matching algorithm, the extracted text data to reference dataset(s) to determine one or more matches between the extracted text data and at least one of the reference dataset(s), wherein the one or more matches are based on at least one similarity score, inputting the determined one or more matches and the at least one similarity score into a trained machine-learning model to refine the one or more matches, and outputting a representation of the refined one or more matches and the at least one similarity score to a graphical user interface of a device.

The invention discloses an animal single cell data cell type annotation method, which is characterized by comprising the following steps: collecting Bulk RNA-seq data of purified cell types in various tissues and organs of a specific animal, and integrating the Bulk RNA-seq data into a reference data set after preprocessing; the method comprises the following steps: acquiring single-cell RNA-seq original data, screening the single-cell RNA-seq original data to obtain high-quality cells, screening high-variation genes from the high-quality cells, processing the high-variation genes, extracting principal components, performing dimensionality reduction on the principal components, and performing cell clustering based on a dimensionality reduction result to obtain cell clusters; and calculating expression similarity between the cell clusters and the reference data set based on the high-variation genes, determining initial cell types of the cell clusters according to a similarity result, and carrying out iterative tuning on the initial cell types with similar scores to obtain a final cell type annotation result. The method provides efficient and accurate technical support for animal single cell research.

Various embodiments of the present disclosure provide prompt engineering and iterative, feedback-based generative techniques that improve traditional LLM technology, including extractive LLM techniques. The techniques may include selecting one or more simple annotated question-answer pairs for an input data object comprising an input question and an input document from a reference dataset. The techniques may include selecting one or more complex annotated question-answer pairs from the reference dataset. The techniques may include generating a few-shot prompt based on the one or more simple annotated question-answer pairs and the one or more complex annotated question-answer pairs. The techniques may include providing the few-shot prompt to a large language model (LLM) to receive a predictive output for the input question.

The invention relates to a computer implemented method for defect detection comprising: obtaining an imaging dataset of an object comprising integrated circuit patterns; obtaining a reference dataset of the object; registering the imaging dataset and the reference dataset by obtaining at least one transformation field pair comprising an input transformation field and a corresponding reference transformation field, wherein the input transformation field or the reference transformation field can be zero; and detecting defects in the imaging dataset using the at least one obtained transformation field pair. The invention also relates to a computer-readable medium, a computer program product and a system for detecting defects.

The invention discloses a building component autonomous inspection system and method based on a BIM (Building Information Modeling) and a VLM (Visual Language Model), and relates to the technical field of intelligent inspection of building engineering. Generating a spatial reference data set, a semantic tag set and a component semantic embedding vector; the method comprises the following steps: collecting multi-modal visual and spatial data, carrying out semantic segmentation and description generation by utilizing a visual language large model, calculating a semantic ambiguity coefficient and carrying out confidence evaluation; through BIM and visual semantic embedding vector comparison and spatial deviation analysis, calculating a cross-modal consistency index and generating a deviation report; calculating a comprehensive path cost coefficient and performing path optimization; and in combination with semantic recognition, cross-modal consistency and path cost, obtaining inspection performance evaluation indexes, realizing adaptive optimization through reinforcement learning, and writing back to a BIM database to realize continuous improvement of data and decision ability.

An example method includes (a) obtaining an object detection from a perception system that describes an object in an environment of the autonomous vehicle; (b) obtaining, from a reference dataset, a label that describes a reference position of the object in the environment; (c) determining a plurality of component divergence values respectively for a plurality of divergence metrics, wherein a respective divergence value characterizes a respective difference between the object detection and the label; (d) providing the plurality of component divergence values to a machine-learned model to generate a score that indicates an aggregate divergence between the object detection and the label, wherein the machine-learned model includes a plurality of learned parameters defining an influence of the plurality of component divergence values on the score; (e) evaluating a quality of a match between the object detection and the label based on the score.

The invention relates to the technical field of riding route planning, and particularly provides a riding route AI planning method based on user demands, which comprises the following steps: receiving user riding preference information and basic travel information, and carrying out integration quantitative preprocessing to form a structured user demand data set; acquiring multi-source reference data such as preset route database data and real-time dynamic data, and performing fusion preprocessing to generate an adaptive reference data set; inputting the two types of data sets into a trained AI planning model to calculate an optimal path, and embedding adaptive point locations based on preference constraints to generate an initial path; and performing multi-dimensional adaptability optimization by combining basic constraints, generating a personalized scheme with multiple versions, setting an output priority, completing route optimization according to a user interaction instruction, and outputting a final executable riding route. The method deeply adapts to multiple demands of the user, improves the planning accuracy and flexibility, and provides a personalized riding route service giving consideration to experience and practicability for the user.

The invention relates to a video generation method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring audio data information and multi-role video frame information; performing dynamic three-dimensional mask processing on the multi-role video frame information to obtain role mask information; combining the audio data information and the role mask information based on multi-modal diffusion to generate role mouth shape action information; acquiring a reference data set, and training a video generation model according to the reference data set based on a preset staged training strategy; and inputting the role mouth shape action information and the multi-role video frame information into the video generation model to generate an effective fusion video. The method and the device can be applied to application scenes of a financial service system and a digital medical system, and can effectively realize video generation with high audio synchronization rate in a multi-role scene according to the input audio data and the multi-role video frame.

The invention discloses a provincial road network emergency resource scheduling path optimization method based on big data analysis, and relates to the field of path optimization, and the method comprises the following steps: obtaining corresponding resource data, and constructing a data storage library according to the obtained resource data; obtaining element information of corresponding resource data in the data storage library, setting a data storage stereoscopic model, and obtaining a corresponding data analysis set; analyzing and processing the data analysis set to obtain importance data between corresponding element information combinations, and setting an element importance evaluation data table; acquiring emergency event information, and setting a reference data set according to the emergency event information; setting basic resource scheduling path information corresponding to the corresponding emergency event information according to the reference data set; performing optimization processing on the basic resource scheduling path information based on the element importance evaluation data table to obtain optimized resource scheduling path information; according to the method, the efficiency and the accuracy of the path optimization process are improved to a certain extent.

An example method includes (a) obtaining an object detection from a perception system that describes an object in an environment of the autonomous vehicle; (b) obtaining, from a reference dataset, a label that describes a reference position of the object in the environment; (c) determining a plurality of component divergence values respectively for a plurality of divergence metrics, wherein a respective divergence value characterizes a respective difference between the object detection and the label; (d) providing the plurality of component divergence values to a machine-learned model to generate a score that indicates an aggregate divergence between the object detection and the label, wherein the machine-learned model includes a plurality of learned parameters defining an influence of the plurality of component divergence values on the score; (e) evaluating a quality of a match between the object detection and the label based on the score.

The present application relates to a terminal device for blood sugar management during pregnancy and its cloud server, the terminal device collects physiological data and blood sugar data of a target user including the gestation period, the cloud server obtains a plurality of blood sugar data of reference users from a plurality of terminal devices and forms an overall reference data set in a manner associated with gestational age, the cloud server calculates basic statistical indicators for each group in the overall reference data set, combines the specific percentile position of the user's blood sugar data in its own group distribution, judges the current blood sugar state of the user by comparing the user's blood sugar data with the preset healthy blood sugar range, and generates a graded diet suggestion accordingly, the diet suggestion optimizes the diet structure of the target user by selecting food categories with different GI values; the terminal device receives and visually displays personalized diet suggestions matching the blood sugar status of the target user to the target user.

This invention relates to a method for designing and utilizing a broad RNA modification library that includes all surrounding sequences of the RNA modification. By introducing multiple types of RNA modifications, this invention enables the construction of RNA modification libraries that encompass all possible combinations of surrounding sequences flanking the modified RNA. Therefore, it differs from existing reported methods for constructing RNA modification libraries. Furthermore, using the RNA modification library of this invention for single-molecule RNA sequencing generates a high-quality reference dataset containing all possible combinatorial motifs. If this dataset is used as training data for deep learning (or machine learning), it overcomes the limitations of existing software that is limited to specific RNA modifications or has low prediction accuracy for other modifications. This allows for the development of novel, more accurate RNA modification detection software and the simultaneous detection of multiple RNA modifications.

The invention provides a method and device for predicting the lifetime of a tumor patient, computing equipment and a storage medium. The method comprises the following steps: acquiring multi-dimensional medical data related to the condition of a tumor patient, wherein the multi-dimensional medical data at least comprises data of at least two dimensions; determining a similar dataset of the multi-dimensional medical data from a reference dataset, the reference dataset including reference multi-dimensional medical data and corresponding lifetime of each of a plurality of reference patients having reached the end of life at different time points; and obtaining a prediction result of the lifetime of the tumor patient based on the multi-dimensional medical data and the similar data set by using a fine-tuned large language model, the prediction result indicating the lifetime of the tumor patient and an explanatory text of information on which the prediction result is based.

The invention relates to the technical field of power battery management, and discloses a power battery generative dataset quality evaluation method, which comprises the following steps: mapping a power battery generative dataset to be evaluated to a standardized feature space to obtain a target generative dataset; performing first data processing on the target generation type data set to obtain a multi-dimensional evaluation result, including: calculating a statistical distribution closeness degree of the target generation type data set and a real reference data set to obtain a authenticity evaluation result; calculating the distribution breadth of the target generation type data set in the feature space to obtain a diversity evaluation result; calculating the multi-variable synchronization relation consistency of the target generation type data set under the same or similar working condition labels to obtain a consistency evaluation result; calculating the coverage integrity of the evaluation target generation type data set on the key working condition area to obtain a representative evaluation result; and performing second data processing on the multi-dimensional evaluation result to obtain a comprehensive evaluation result. According to the method, the power battery generation type data quality can be measured in multiple dimensions.

The embodiment of the invention relates to the technical field of wind measurement data processing, and provides a wind measurement data correction method and device, electronic equipment and a medium, and the method comprises the steps: obtaining equipment wind measurement data which is first time period measurement data containing a time sequence, and obtaining second time period wind data of a position where the equipment is located; extracting a first reference data set, a second reference data set and a third reference data set from the first time period measurement data and the second time period wind data; respectively carrying out matrix processing on the first reference data set, the second reference data set and the third reference data set to obtain a first wind frequency matrix, a second wind frequency matrix and a third wind frequency matrix; and determining a correction result of the wind measurement data based on the first wind frequency matrix, the second wind frequency matrix and the third wind frequency matrix. And high-precision data correction comprehensively considering two-dimensional characteristics of wind speed and wind direction is realized.

The invention provides a heat exchange tube folding control method and system based on folding state analysis, and the method comprises the steps: obtaining the real-time folding state information and stress information of a heat exchange tube through various types of sensors, and generating complete data through preprocessing operation; determining a plurality of influence factors influencing the folding of the heat exchange tube, constructing a heat exchange tube folding enthusiasm prediction model, and obtaining an optimal factor group of the folding of the heat exchange tube through the enthusiasm prediction model; calling reference data related to the optimal factor group from a historical database, respectively generating a to-be-adjusted data set and a reference data set, and calculating a similarity value of the two data sets; and according to a comparison result of the similarity value and a preset threshold value, power source output during folding of the heat exchange tube is adjusted, and the folding state is controlled. According to the method, the similarity between the current state and the historical optimal state of the heat exchange tube is calculated by utilizing the optimal factor group, so that the folding of the heat exchange tube is accurately controlled, and the folding process of the heat exchange tube is optimized.

The application discloses a data processing method, which comprises the following steps: assembling an attribute set by referring to the attributes of the data of each category in a reference data set; screening a candidate attribute subset from the attribute set; wherein the candidate attribute subset is a subset in the attribute set that satisfies a mutual information condition with a value domain set composed of the values of the data of each category; analyzing the association rules between each attribute subset in the candidate attribute subset based on the classification identifiers of each category, and locating a frequent item set with a target association rule; wherein the support degree of each item set in the frequent item set with the target association rule is greater than a support degree threshold, and the confidence degree of each item set is greater than a confidence degree threshold, and each item set comprises at least two attribute subsets. The application also discloses a data processing device, an electronic device and a computer readable storage medium.

The application relates to a special variable user load feature extraction method, device, equipment and medium based on multi-dimensional time sequence data. The method comprises the following steps: preprocessing an original data set collected to obtain a reference data set, obtaining a to-be-repaired data point in the reference data set, respectively obtaining a spatial repair value and a time repair value of the to-be-repaired data point, fusing the spatial repair value and the time repair value based on a fusion weight to obtain a data repair value, updating a data value corresponding to the to-be-repaired data to the data repair value to obtain a repaired data set, performing time sequence alignment on a load repaired data set and an environment repaired data set to obtain a multi-dimensional time sequence data set, extracting a time sequence feature set of the multi-dimensional time sequence data set, screening the time sequence feature set to obtain a target feature set, and constructing a special variable user load space-time feature sample library based on the target feature set. The method can improve data processing precision.

The invention relates to methods for providing one or more reference data sets for navigation of a mobile device (200) that is moving or is intended to move in an environment, comprising: providing boundary information for a boundary (223) that at least partially limits a working area (222) for the mobile device in the environment, wherein the boundary has been determined based on environmental and / or device information, wherein the environmental and / or device information is or has been obtained by means of at least one sensor of the mobile device when the mobile device moves in the environment, in a first step; causing the mobile device to move along the boundary and / or within the working area based on the boundary information, in particular automatically, in a second step, and to acquire one or more reference images by means of the at least one sensor;Generating one or more reference datasets, wherein one or each of the multiple reference datasets contains one of the captured reference images and a reference pose (244b) of the mobile device at the time the respective reference image was captured; and providing the one or more reference datasets for correcting a pose of the mobile device when the mobile device is moved.

Embodiments of the present disclosure relate to a method, an electronic device, and a computer program product for dataset updating. The method includes: acquiring a reference dataset comprising a plurality of reference images, wherein each reference image has a soft label, and the soft label indicates a probability that the each reference image belongs to each target class in a plurality of target classes. The method may further include: determining, according to a similarity between an input image received and the each reference image, an image soft label for the input image. The method further includes: adding the input image and the image soft label for the input image to the reference dataset to update the reference dataset.

The present application relates to the technical field of dynamic evaluation of spinal morphology, and particularly relates to a child spinal morphology intelligent evaluation method and system based on big data analysis. The method comprises the following steps: S1: obtaining static data and dynamic data, dividing the static data into a high-priority evaluation data set and a basic reference data set, and dividing the dynamic data into a key change data set and an auxiliary monitoring data set; taking the complete time period between the time point when the child spinal morphology first receives medical evaluation and the time point when the system confirms that the child spinal function and morphology reach a steady state as an evaluation and diagnosis cycle, and distributing the static data and the dynamic data to the evaluation and diagnosis cycle to obtain a data-stage distribution result. The present application constructs an intelligent evaluation system, accurately tracks child spinal development, dynamically optimizes data collection and stage distribution through error analysis and similarity comparison, and realizes individualized, efficient and self-adaptive closed-loop diagnosis management.

Methods, apparatus, systems, and articles of manufacture to estimate cardinality through ordered statistics are disclosed. In an example, an apparatus includes processor circuitry to selects a sample dataset from a first reference dataset of media assets and partitions the sample dataset into m mutually exclusive subsets of approximately equal size. The processor circuitry then estimates a ratio of a sample weighted average and empirical cumulative distribution of an approximately largest order statistic from at least one of the m subsets and generates an estimate of a total cardinality of the first reference dataset by multiplying the ratio by approximately m.

Embodiments of the present application disclose a parameter configuration method, device, equipment, storage medium and product. The method comprises: obtaining a shooting parameter of a target camera device and a reference data set; determining a target virtual scene parameter corresponding to virtual content to be shot by the target camera device according to the shooting parameter of the target camera device and the reference data set; displaying the virtual content according to the target virtual scene parameter; and enabling the target camera device to shoot the virtual content. It can be seen that the target virtual scene parameter corresponding to the virtual content to be shot by the target camera device is determined by using the reference data set, which can improve the configuration efficiency and accuracy of the virtual scene parameter.

The application discloses a data processing method, device and computer readable storage medium. The method comprises the following steps: obtaining a first data set generated by a target object in a first time window; predicting a second data set corresponding to a second time window according to fluctuation amplitudes between data amounts corresponding to N time points; obtaining a first reference data set corresponding to a first historical time window; predicting a third data set corresponding to a third time window according to the first data set, the second data set and the first reference data set; and generating a predicted data set of the target object according to the second data set and the third data set. The application can improve the accuracy of the predicted data set.

Disclosed is wearable device for monitoring brain activity and for providing non-invasive stimulation, wearable device comprising: headwear arrangement comprising: electrode arrangement comprising electrodes configured to make electrical contact with skin of user; and flexible cushioning member supporting electrodes; input / output (I / O) arrangement configured to: receive electrical signals (ES) from electrode(s); transmit ES to data processing arrangement (DPA) to process ES for generating brain stimuli (BS); and generate and apply BS generated using a brain stimulation protocol (BSP) to electrode(s), BS applied on skin lying partially in mastoid region; DPA comprising processing unit configured to: receive ES from I / O arrangement; analyse ES with predetermined reference data set, by extracting signal feature(s); apply processing algorithm(s), to map signal feature(s) to stimulation parameter(s); generate BSP comprising stimulation parameter(s); and transmit BSP to I / O arrangement; and power unit(s) supplying electrical power to I / O arrangement and DPA.

The invention discloses an operation positioning management system and method for side channel tracking, and relates to the technical field of big data analysis, and the method comprises the steps: collecting the historical power consumption and electromagnetic signal data of normal operation of target equipment, adding a timestamp, and carrying out the preprocessing to form a reference data set; constructing a data standard library in combination with operation codes and timestamps extracted from the operation logs; comparing the real-time power consumption, the electromagnetic signal data and the operation code with a standard library, and positioning an attack starting node and a key operation node; quantitatively evaluating a node risk level; and generating a report list according to the risk level and dynamically tracking. The system comprises a reference data construction module, a data standard library construction module, an operation node positioning module, a risk level evaluation module and a dynamic tracking management module which are electrically connected in sequence. According to the method, accurate positioning, risk quantitative evaluation and dynamic tracking of attack nodes are realized, and clear targeting is provided for protection.

The invention discloses a data supervision system and method for targeted heavy metal removal in a fly ash washing process, and relates to the technical field of fly ash resource.The method comprises the steps that the heavy metal type, concentration, treatment object volume and period duration between two times of verification are periodically collected, the heavy metal content is calculated, and an accumulation table is generated; after data verification, grouping according to cycle duration, checking the average value of the heavy metal content, and constructing a standard reference table; determining a heavy metal conversion relation through deviation value analysis, and quantitatively generating a reference data set; and monitoring the sudden change of the concentration in real time, comparing with a reference value after conversion, and triggering verification early warning or process verification according to a threshold value. The system comprises a periodic data acquisition module, a heavy metal content accounting module, a standard reference table construction module, a reference value quantitative generation module and an abnormity monitoring and handling module, and the modules are in closed-loop connection. According to the method, full-process automatic supervision is realized, the abnormal response efficiency and the supervision accuracy are improved, the labor cost is reduced, and compliance guarantee is provided for resource utilization of the fly ash.

A method and a device for monitoring a field of view of a stationary sensor. The method includes: a first step for receiving recordings of surroundings of the sensor, which are generated at different times within a predefined measuring time frame with the aid of the sensor; a second step for ascertaining, based on the recordings, measuring points, which each represent positions and / or dimensions of objects in the surroundings of the sensor; a third step for grouping the ascertained measuring points based on a predefined similarity metric, where each group of measuring points is represented by a prototype; and a fourth step for ascertaining a change of content within the field of view of the sensor based on a comparison of the prototypes with a reference data set, which includes reference prototypes, whose underlying measuring points are ascertained in a reference state of the sensor and of the surroundings.

The present disclosure proposes a calibration method and system for a nacelle-type laser radar. In an area with flat terrain, a wind tower and a placement tower are set at a preset distance apart. A sensor group is arranged on the wind tower, and a test platform, an engine, and a nacelle-type laser radar are arranged on the placement tower. The nacelle-type laser radar is set on the test platform, and the engine is located below the test platform, with the engine bracket in contact with the test platform. The calibration method includes: while the engine is running, using the nacelle-type laser radar and the sensor group to collect data respectively to obtain a measured data set and a reference data set; calculating a slope and a correlation coefficient based on the measured data set and the reference data set; if the slope and correlation coefficient meet the requirements, calculating the uncertainty based on the measured data set and the reference data set, and calibrating the nacelle-type laser radar based on the uncertainty. The method according to the present disclosure can improve the accuracy of nacelle-type laser radar calibration.