96 results about "Dimensionality reduction algorithm" patented technology

The invention relates to the technical field of textile detection, in particular to a computer-aided textile color fastness prediction method, which comprises the following steps: acquiring multi-source data before and after a textile color fastness test through computer control equipment, including digital images, reflection spectrum data, process parameters and dye characteristic component data, carrying out pretreatment; extracting color difference features and spectrum similarity features, and screening by adopting a dimension reduction algorithm to form a core feature vector; a machine learning algorithm is adopted to construct a color fastness prediction model, parameters are optimized in combination with cross validation and an early stop mechanism, model fine tuning is executed, and a final optimization model is obtained; and outputting a color fastness grade prediction result, and carrying out spectrum similarity secondary verification. According to the method, objective, efficient and high-precision prediction of the color fastness grade of the textile is realized, and the problems of high subjectivity and low efficiency of traditional manual evaluation are solved.

The invention relates to the technical field of natural language processing, in particular to a network public opinion hot topic recognition method, device and equipment, and the network public opinion hot topic recognition method comprises the steps: embedding social media text data through a Chinese pre-training language model to obtain corresponding high-dimensional vector representation; performing dimension reduction on the high-dimensional vector representation by adopting a dimension reduction algorithm based on manifold learning to obtain dimension-reduced text data; and performing density clustering on the text data subjected to dimension reduction, and extracting keywords from a clustering result as recognized topics. According to the method, public opinion topics with high attention and representativeness can be accurately recognized and summarized from massive social media data, the automation level of public opinion hotspot recognition is improved, and more accurate and timely information support is provided for public opinion analysis and decision making.

The invention discloses a hydraulic motor abrasion monitoring method and system, and belongs to the technical field of abrasion fault prediction.The method specifically comprises the steps that multi-dimensional data signals of a hydraulic motor are collected in real time and preprocessed; the preprocessed multi-dimensional data of the hydraulic motor are processed, the early wear characteristics of the hydraulic motor are extracted, and the processing comprises decomposition of the preprocessed multi-dimensional data of the hydraulic motor, construction of a decomposition tree, reservation of nodes with preset frequency in the decomposition tree, and reconstruction, difference calculation and window moving average suppression of the reserved nodes; fusing the early wear characteristics of the hydraulic motor, and generating a comprehensive health index through a dimension reduction algorithm; judging whether the hydraulic motor has early-stage fine wear or not according to the comparison of the comprehensive health index and a dynamic threshold value, and giving an alarm according to a judgment result; according to the method, early-stage fine wear signs of the hydraulic motor can be found, a maintenance plan is made in advance, and production interruption caused by sudden equipment failures is avoided.

The invention provides a power battery feature extraction method and system, electronic equipment and a storage medium, and the method comprises the steps: obtaining the operation data of a power battery, including time-varying electrical data and thermal data; according to the operation data, multi-modal features at multiple time points are extracted, a multi-modal feature matrix is obtained, and the multi-modal features comprise electrical features and thermal features; analyzing time sequence correlation among different features in the multi-modal feature matrix, constructing a time sequence correlation matrix, performing feature decomposition on the time sequence correlation matrix, and constructing a dimensionality reduction projection matrix; and performing dimension reduction processing on the multi-modal feature matrix according to the dimension reduction projection matrix to obtain a target feature matrix, and completing feature extraction. Through collaborative design of multi-modal feature extraction and a dimension reduction algorithm based on time sequence perception, the precision and efficiency of power battery feature extraction are effectively improved.

An efficient open set encryption traffic identification method for super-large scale traffic relates to the technical field of data security, solves the technical problems of insufficient resource efficiency, feature expression and open set identification in the prior art, and comprises the following steps: S1, inputting a network traffic data set, and generating dimension reduction data of an optimal dimension through a PCA dimension reduction algorithm; s2, inputting the dimension reduction data of the optimal dimension into an improved MobileNetV3 model, and in combination with an improved SE attention mechanism and a composite structure of Dense connection and residual connection, extracting features and outputting a classification result; s3, judging whether a classification result belongs to a known category, a drift sample or an unknown category by using an open set recognition algorithm based on confidence; s4, if the classification result is a drift sample or an unknown category, incremental learning based on knowledge distillation is started, parameters of the improved MobileNetV3 model are updated, and old knowledge is reserved. According to the invention, the network security protection capability can be effectively improved.

The invention discloses a hydrogen fuel cell life prediction method and system, and relates to the field of new energy, and the method comprises the steps: compressing a feature dimension through employing a dimension reduction algorithm, and obtaining a feature vector sequence reflecting the degradation state of a hydrogen fuel cell; constructing a state-of-health index reflecting the performance degradation degree of the hydrogen fuel cell, and constructing a state-of-health curve in combination with the time sequence; based on the change trend of the health state curve, dividing the life process of the hydrogen fuel cell into an early-stage stable stage, a middle-stage slow degradation stage and a later-stage rapid degradation stage by adopting a segmented modeling algorithm; based on the health state index and the feature vector sequence, constructing a time sequence prediction model, and predicting the residual service life of the fuel hydrogen fuel cell; and monitoring a sudden change behavior of the health state curve, and identifying an unexpected degradation condition. According to the method, feature extraction and dimension reduction processing are carried out on the operation data of the hydrogen fuel cell, the health state index and curve are constructed, and the prediction precision of the remaining service life is remarkably improved.

The invention provides a dynamic deformation monitoring data real-time processing and analysis method, and relates to the technical field of data processing. The dynamic deformation monitoring data real-time processing and analysis method specifically comprises the following steps: S1, deploying a multi-type sensor network, collecting displacement, acceleration and inclination angle data in real time, and performing time synchronization and data preprocessing through a wireless / wired transmission protocol; s2, mapping the multi-dimensional monitoring data into a high-dimensional space-time trajectory, and generating a visual trajectory diagram by adopting a dimension reduction algorithm; s3, segmenting the track based on a change point detection algorithm, and extracting curvature, speed and acceleration change rate characteristics of each segment; s4, constructing a dynamic behavior pattern library, and performing real-time pattern matching and switching detection in combination with a hidden Markov model and a support vector machine; and S5, identifying abnormal points through a 3 sigma criterion and a multi-model voting mechanism, evaluating a risk grade according to a mode switching frequency, and triggering graded early warning. According to the invention, multi-source cooperation is realized, and early warning precision is high.

The invention relates to the technical field of computer vision, in particular to a fabric color fastness analysis method and system based on image processing. Comprising the following steps: acquiring a multispectral image of a to-be-detected fabric, and synchronously acquiring multi-factor data; processing the multispectral image to generate an enhanced image; performing image recognition on the gray sample card in the enhanced image to generate a calibration image; performing spectral data projection on the calibration image through a dimension reduction algorithm, and performing color analysis and quantification on color change to generate color difference features; extracting texture features by using a conditional generative adversarial network, carrying out image recognition and calculation on the calibration image through a decoupling algorithm, and generating a region credibility graph; and inputting the chromatic aberration features, the texture features, the multi-factor data and the regional credibility map into an image feature fusion model for mapping, performing analysis through gradient visualization, and outputting an abnormal feature saliency map. According to the method, a color fastness objective analysis closed loop is created, so that the accuracy and the universality of an analysis result are improved.

The invention discloses an image management method and device. The method comprises the following steps: in response to an operation of uploading a target image to a cloud disk, performing feature dimension reduction extraction on the target image by using a preset feature dimension reduction algorithm to obtain a target feature set; performing label identification on the target image according to the target feature set by using a preset label identification algorithm to obtain a plurality of labels corresponding to the target image; performing semantic fusion on the plurality of tags by using a preset tag fusion algorithm to obtain a target tag; and in the cloud disk, associatively storing the target label and the target image. According to the method, the target image uploaded by the user is intelligently processed through feature dimension reduction, tag identification and semantic fusion algorithms, the target tag with refined semantics is generated, and the target tag and the image are associated and stored in the cloud disk, so that the management degree of the image in the cloud disk is improved.

The present invention is in the field of geological measurement, and particularly relates to a method and system for intelligent removal of multi-dimensional logging data outliers, aiming to solve the problem that it is difficult to achieve stable parameter acquisition in the drilling process when there are more outliers in the prior art. The present invention comprises: acquiring a multi-dimensional logging curve; obtaining two-dimensional logging curve data by using the Umap dimensionality reduction algorithm to perform dimensionality reduction; establishing a first data point index of the multi-dimensional logging curve and the two-dimensional logging curve data; obtaining valid logging data by processing outliers with an isolation forest algorithm based on the two-dimensional logging curve data; acquiring a second data point index based on the first data point index and valid logging data; finding marked data points where the second data point index exists in the multi-dimensional logging curve data.

The invention discloses a cement-based material phase identification and quantitative analysis method combining artificial intelligence and expert knowledge. The method comprises the following steps: acquiring a BSE image, a qualitative element surface spectrum and a quantitative element surface spectrum of a cement-based cementing material sample; synthesizing a color EDS image based on the qualitative element surface spectrum, and performing superpixel division on the image; based on the quantitative element surface spectrum, extracting an average value of element relative contents of all pixels in each superpixel range as an element feature; performing two-dimensional visualization processing on all the element features by adopting a PHATE dimension reduction algorithm to generate derivative element features; performing manual correction and phase label labeling on a clustering result by utilizing a Glue multi-view interaction platform in combination with expert priori knowledge; and fusing the corrected clustering result with the superpixel division result, and regenerating a new phase classification mask with a phase identifier. According to the method, intelligent identification and high-precision analysis of the cement-based material multiphase system can be realized, and the accuracy, reliability and interpretability of phase classification are improved.

The invention belongs to the technical field of food safety detection, and particularly relates to a method for detecting polluted vegetable oil by using a deep learning assisted three-dimensional fluorescence spectrum. According to the method, firstly, the optimal adsorption condition of magnesium silicate to the vegetable oil is predicted based on a linear regression model and an L-BFGS-B algorithm, and targeted separation of mineral oil in the polluted vegetable oil is achieved through adsorption of magnesium silicate to the vegetable oil under the optimal condition. Then, on the basis of mineral oil pollutant samples obtained through separation, characteristic fingerprint spectrums of different mineral oil types are extracted from the collected three-dimensional fluorescence spectrums by applying multiple pre-training network models, and therefore several common mineral oil pollutant types are distinguished. And decomposing and judging fluorescent components of corresponding pollutants in the mineral oil by combining a parallel factor data dimension reduction algorithm. And finally, building a support vector regression model of the corresponding component based on the corresponding relationship between the fluorescence signal and the concentration of each pollution component, thereby realizing the quantitative detection of each pollution component in the mineral oil.

A system, method, and computer-program product includes receiving a plurality of vector embeddings having an initial dimensionality and projecting the plurality of vector embeddings into lower-dimensional spaces using at least two different dimension reduction algorithms to generate corresponding sets of projected vector embeddings. Each set of projected embeddings may be quantized and nearest neighbors for the original embeddings and for each quantized set of projected embeddings may be calculated. Additionally, a neighbor preservation metric may be evaluated for each quantized set by comparing its nearest neighbors to those of the original embeddings. Based on the neighbor preservation metrics and a predefined error tolerance, an optimal compression configuration may be selected.

The invention discloses a sea surface small target detection method and system based on a magnetic KNN and an optimized concave packet, and belongs to the technical field of radar signal processing. Comprising the following steps: extracting high-dimensional features of sea clutters and echoes to be detected, and selecting three-dimensional feature points of the clutters and the echoes to be detected according to a dimension reduction algorithm; processing the three-dimensional feature points of the clutters by adopting a constant false alarm algorithm to obtain clutter feature points after constant false alarm; on the basis of the clutter feature points after the constant false alarm, constructing a concave packet by adopting an optimized concave packet construction algorithm; the clutter feature points in the concave packet serve as vertexes of a tetrahedron, and the concave packet is divided into tetrahedral grids; positioning the points in the tetrahedrons by using a pointLocation function, judging the indexes of the tetrahedrons where the points are located, and if the indexes exist, judging that the points are clutters; and if the index does not exist, judging as a target. According to the detection method, the construction efficiency of the concave packet is also improved while the relatively high detection probability is ensured, and the method has relatively good generalization.

This invention relates to the field of wind power generation technology and discloses a method, device, equipment, and medium for fault assessment of wind turbine transmission systems. The method utilizes a data dimensionality reduction algorithm to reduce the dimensionality of multidimensional raw data, retaining core distinguishing features and simplifying calculations. Then, a data clustering algorithm is used to intelligently classify operating states, and a preliminary fault mode mapping is constructed by combining historical faults. Subsequently, the rationality of clustering is verified using signal source correlation coefficients, and long-term historical operating data is filtered through historical data matching rates to reduce the risk of misjudgment. Next, based on time series analysis and degradation path analysis, the coupling relationship between vibration trend slope, temperature accumulation offset, and torque decay cycle is obtained. Finally, core features are extracted through convolutional neural networks to accurately output the probability distribution and specific location of fault occurrence, thereby improving the accuracy of fault identification in the transmission system of offshore wind turbines and the ability to predict component performance degradation.

The invention discloses an identity verification method based on touch sliding gesture features in combination with AI and biological feature recognition technologies. According to the technology, multi-dimensional features of user touch behaviors are extracted through feature engineering, a full-connection neural network is innovatively adopted to deeply mine a data relationship, and a manifold learning theory and multiple dimension reduction algorithms are applied to optimize a data processing flow. In addition, a variational auto-encoder (VAE) is introduced to generate non-owner user data, so that the classification precision and generalization ability of the model are effectively enhanced. Compared with a traditional biological recognition technology (such as fingerprints, faces and irises) and behavior biological feature recognition (such as gaits and keyboard tapping rhythms), the method does not need additional hardware, is higher in privacy protection and is suitable for existing intelligent equipment. Through deep algorithm research and software testing, the system obtains a certain achievement, and an innovative scheme is provided for identity safety verification of intelligent equipment.

The invention discloses a data analysis method and system based on questionnaire investigation, and particularly relates to the technical field of data analysis, and the method comprises the following steps: S1, collecting and preprocessing questionnaire data, and obtaining original questionnaire data; s2, questionnaire feature engineering: extracting core features from the standardized questionnaire data set based on a joint screening mechanism of semantic similarity and information gain, and obtaining low-dimensional feature vectors through a dimension reduction algorithm; s3, subject mining and user group division are carried out on the low-dimensional feature vectors; and S4, outputting a result based on the analysis model, and generating a targeted questionnaire analysis report which comprises topic distribution, group difference and potential demand association conclusions. According to the data analysis method and system based on questionnaire investigation, through a multi-strategy fusion preprocessing mechanism, a combined screening feature engineering method, a fusion modeling analysis strategy and a self-adaptive model optimization module, the accuracy and efficiency of questionnaire data analysis are improved.

This invention relates to a big data topic analysis method based on an embedding model. First, the Sentence-BERT model is used to perform sentence embedding representation on preprocessed Chinese text data. Then, the UMAP projection dimensionality reduction algorithm is used to reduce the dimensionality of the embedded vectors. Next, the HDBSCAN clustering algorithm is used to cluster the dimensionality-reduced vectors. Based on the assignment of each Chinese text in the target Chinese dataset to a corresponding topic class, the Chinese words with the highest c-TF-IDF scores are selected to represent each topic class. Finally, the DSG model is used to perform word embedding representation on the topic words, calculating the similarity between different topic words and between different topic classes, thereby detecting the volatility of newly emerging topic classes. The entire scheme design has higher topic consistency and topic diversity, and can detect new hot topics in a timely and accurate manner, providing early warnings.

The invention discloses a multi-label smell description prediction method, and relates to the field of compound smell prediction, and the method comprises the steps: obtaining compound identification information, molecular structure descriptors and smell label data, and constructing a multi-label smell data set; generating a molecular structure feature vector through a molecular fingerprint coding technology, and extracting a multi-dimensional descriptor reflecting the physicochemical properties of molecules; compressing the molecular fingerprint features to a low-dimensional space through a dimension reduction algorithm; performing unbalanced data processing on the training set, fusing the dimension-reduced molecular fingerprints with the molecular descriptors to form a joint feature matrix, and configuring a class weight balance mechanism and overfitting suppression parameters by adopting a multi-label classification architecture; independently optimizing a probability threshold for each odor label based on the verification set; and outputting a multi-odor label combination prediction result according to the target molecule identification information. According to the scheme, the multi-odor characteristics of the compound can be accurately depicted, and the combined recognition accuracy of the compound odor is remarkably improved.

A system and method of determining efficacy of treatment by at least one processor may include receiving, from at least one camera, images depicting motion of an animal that may be treated with a predetermined substance of interest. Said processor may extract from the images, a plurality of motion features representing motion of at least one specific body part of the animal, and apply a dimensionality reduction algorithm on the plurality of motion features, to obtain a latent vector representing the plurality of motion features in a latent space. The latent vector may include a plurality of latent features. Said processor may subsequently calculate a value of a behavioral indicator, representing a behavior of the animal, based on the latent features of the latent vector, and determine efficacy of the treatment based on the behavioral indicator value.

A large language model-driven unit combination optimization method, relating to the field of power system technology, includes: acquiring basic economic operation data of the power system; constructing and solving a unit combination model with integer relaxed security constraints based on the basic economic operation data; obtaining and solving the unit combination model; constructing variable dimensionality reduction algorithms, constraint dimensionality reduction algorithms, and infeasibility solution repair algorithms; using a large language model to perform single-algorithm evolution on the variable and constraint dimensionality reduction algorithms; using a large language model to perform single-algorithm evolution on the infeasibility solution repair algorithm; solving and calculating the unit combination model with integer relaxed security constraints; calculating the variable constraint dimensionality reduction security constraint unit combination model and obtaining the optimal solution; and using the optimal infeasibility solution repair algorithm to repair the optimal solution of the variable constraint dimensionality reduction security constraint unit combination model, obtaining the result. This method addresses the shortcomings of existing model dimensionality reduction methods in terms of generality, stability, and interpretability.

The invention relates to the technical field of textile industry automation and computer vision detection. According to the lint cotton foreign fiber cleaning and detecting system based on visual identification, a hyperspectral imager is used for conducting multi-band scanning on cotton flow, and standardized spectral data are generated; constructing a three-dimensional tensor, and dividing the three-dimensional tensor into a plurality of sub-tensors according to a spatial dimension; carrying out truncated high-order singular value decomposition processing on each sub-tensor, retaining a principal component of a core tensor and removing a secondary component, and generating a sub-tensor after dimension reduction; splicing the sub-tensors after dimension reduction into a global feature map according to spatial positions, and generating candidate foreign fiber masks; triggering a removing device to separate foreign fibers according to the candidate foreign fiber mask, and dynamically adjusting the partitioning size and dimension reduction parameters so as to solve the problems that a traditional dimension reduction algorithm is high in processing delay and cannot meet the requirement of high-speed cotton flow; the signal-to-noise ratio is low and the false detection rate is high due to lack of dynamic feature screening; the problems that static parameters cannot be dynamically optimized, multi-machine cooperation performance is degraded, and classification precision is limited are solved.

The invention belongs to the technical field of oil field logging, and discloses an oil field logging data optimization method based on machine learning and a related device. Original oil field logging data are collected through multi-source logging equipment and preprocessed to generate a preprocessed data set; feature correlation analysis and model importance evaluation are carried out on the data set to extract a key feature set through a dimension reduction algorithm, then the key feature set is adopted to train a pre-constructed machine learning model to obtain a logging data optimization model, and finally the model is applied to optimize to-be-processed logging data and output an optimization result. By adopting the method, the effectiveness and representativeness of logging data characteristics can be remarkably improved, the real condition of an underground reservoir can be accurately reflected, the full-well-section high-resolution logging data can be simply, conveniently and quickly obtained at low cost, a reliable basis is provided for a shale oil and gas development scheme, and the rationality of the development scheme and the overall development efficiency are further enhanced.

The invention relates to the technical field of performance detection, and discloses a metering equipment full-performance detection result automatic determination method, equipment and a medium, and the method comprises the steps: collecting multi-dimensional performance parameters of metering equipment under different test conditions, carrying out the preprocessing and preliminary feature extraction of the multi-dimensional performance parameters, and carrying out the calculation of the full-performance detection result of the metering equipment; performing intelligent comprehensive processing by using a nonlinear high-dimensional dimension reduction algorithm to obtain comprehensive performance parameter data; the comprehensive performance parameter data is intelligently detected through a detection algorithm based on machine learning, a detection result is obtained, a judgment result is automatically generated based on the detection result in combination with self-adaptive threshold adjustment, and the method can solve the technical problems that equipment performance parameters are not comprehensively considered, processing is not accurate, and detection and judgment are not accurate.

The present invention discloses a method, system and device for flame detection of surveillance videos. The method includes: obtaining video data, and determining a candidate region set after preprocessing; extracting abstract features and high-dimensional histogram of oriented gradients (HOG) features from the candidate region set; finding the best candidate region from the candidate region set according to a logistic regression observation model, and determining the target position in the best candidate region; performing dimensionality reduction processing on the abstract features and high-dimensional HOG features of the target position according to a dimensionality reduction algorithm, and determining the weight of each-dimensional feature vector according to the feature contribution degree; and determining whether the target position is a flame according to the Cayley-Klein metric learning criterion and each-dimensional feature vector. In the embodiments of the present invention, flame detection is performed on video data through a machine learning method and a neural network algorithm, so as to solve the problem that traditional manual detection and identification of fires may lead to a more serious fire situation due to negligence.

This invention discloses a rapeseed quality detection method based on feature wavelength optimization and multi-model fusion. The method involves collecting full-band near-infrared diffuse reflectance spectral data of rapeseed samples and simultaneously measuring the true values ​​of physicochemical indicators to construct an original dataset. Based on the data distribution characteristics of each physicochemical indicator, the original dataset is divided into a training set and a test set. Preprocessing algorithms are selected for each physicochemical indicator to process the full-band near-infrared diffuse reflectance spectral data. Dimensionality reduction algorithms are selected for each physicochemical indicator to extract feature wavelengths from the preprocessed spectral data. Predictive models for each physicochemical indicator are established based on the training set, and the performance of the predictive models is verified using the test set to determine the optimal algorithm combination for each physicochemical indicator. The variable importance projection algorithm is used to calculate the comprehensive contribution score of each wavelength point in the full-band spectrum to the four physicochemical indicators. A multi-indicator comprehensive threshold is set, and the core light source wavelength of the portable device is determined from the high-score region.

Provided is a method, system, and computer program product for performing automated feature dimensionality reduction without accuracy loss. A processor may determine a first training value associated with a first dataset of a machine learning model. The processor may rank features of the first dataset in relation to the first training value. The processor may compare the ranked features of the first dataset to a predetermined threshold. The processor may generate a second dataset from the first dataset by removing a third dataset, the third dataset having a set of features that did not meet the predetermined threshold. The processor may determine a second training value associated with the second dataset. The processor may compare the first training value to the second training value. In response to the second training value being lower than the first training value, the processor may analyze the third dataset with a dimensionality reduction algorithm.

The invention discloses an unsupervised dimensionality reduction visualization method for cell image data. The method comprises the following steps: acquiring an unlabeled cell image high-dimensional data set; processing the high-dimensional data set by adopting an unsupervised dimension reduction algorithm to generate low-dimensional embedding representation; generating a first visual chart based on the low-dimensional embedded representation to display the distribution of the data in the low-dimensional space; performing unsupervised clustering analysis on the data points in the low-dimensional embedding representation, and identifying at least one clustering center point; generating a second visual chart based on a clustering analysis result, and marking a clustering center point in an identifiable manner; according to the method, the clustering center point can be accurately identified, and a visual data distribution overview is provided for a user.

The invention discloses an intelligent power distribution system optimization control method based on a knowledge graph, and the method comprises the steps: collecting and preprocessing multi-source power distribution operation data, and constructing a power distribution knowledge graph; performing feature extraction and fusion to generate a power distribution high-dimensional state matrix; constructing a neighborhood graph and executing an Isomap manifold learning algorithm to generate a low-dimensional manifold space; constructing an improved Brown bridge diffusion model, and generating a state evolution path; multi-class risk probabilities are calculated, and a power distribution state risk field is generated; and executing a sequential quadratic programming algorithm, and solving an optimal control strategy. By introducing knowledge graph semantic expression, a curvature correction manifold dimension reduction algorithm, an improved Brownian bridge diffusion model and a sequential quadratic programming algorithm, accurate prediction, risk perception and optimal control of the operation state of the power distribution system are realized.

The present invention relates to the field of data processing and visualization, and specifically to a method, device, and medium for dimensionality reduction and visualization of high-dimensional scientific and technological innovation data. The method includes the following steps: obtaining high-dimensional scientific and technological innovation data, performing data cleaning and standardization processing to obtain a preprocessed data set; extracting features from the preprocessed data set and automatically selecting a dimensionality reduction algorithm to generate dimensionality reduction parameters; performing distributed dimensionality reduction processing based on the dimensionality reduction parameters to obtain a low-dimensional data representation; inputting the low-dimensional data into an intelligent interactive visualization tool for three-dimensional visualization display, and dynamically optimizing it in combination with the user's interaction behavior. The present invention improves the processing efficiency of high-dimensional data and supports users' in-depth interaction and analysis of data.