62 results about "Subspace clustering" patented technology

The invention relates to the technical field of e-commerce supervision and prediction, in particular to a cross-border e-commerce retail risk supervision method and system. Comprising the following steps: acquiring cross-border e-commerce multi-source e-commerce data in real time by using the Internet of Things and an API interface; identifying heterogeneous risk factors by using a deep belief migration network; constructing a risk evolution model and generating a dynamic graph based on the spatial state transition type Markov convolutional network; adopting an abnormal subspace clustering algorithm to carry out self-adaptive clustering and real-time grading early warning on the risk modes; and finally, combining a reinforcement learning algorithm, an intelligent decision and a dynamic optimization risk intervention strategy with resource allocation, and outputting a whole-process risk management and control scheme. The intelligent, automatic and precise levels of cross-border e-commerce retail risk monitoring, early warning and response are improved, risk disposal lag and resource waste are effectively reduced, and the industry supervision safety guarantee capability is enhanced.

The invention discloses a three-dimensional scene reconstruction and monitoring method based on multi-view fusion and deep learning. The method comprises the following steps: acquiring image data from different viewpoints through a plurality of cameras, and carrying out geometric calibration; denoising, correcting and enhancing the image; constructing a multi-scale convolutional neural network and a variational auto-encoder model, and extracting multi-level features; carrying out weighted fusion on the features, and carrying out three-dimensional reconstruction through sparse coding and a graph neural network; performing feature optimization by applying a dynamic graph convolutional network and a double attention mechanism, and performing model updating based on adversarial gradient descent; anomaly detection is carried out through multi-scale analysis and an adversarial variational auto-encoder, preliminary processing is carried out at a camera end by adopting an edge computing technology, and further analysis is carried out at a central server end through a heterogeneous graph neural network and sparse subspace clustering. According to the method, high-precision and intelligent three-dimensional scene reconstruction and monitoring are realized, and the method has a wide application prospect.

The invention relates to a low-dimensional subspace clustering method based on projection matrix guidance, and the method comprises the steps: extracting a light response non-uniformity PRNU noise residual error from input image data through employing a denoising filter, and constructing a PRNU feature data set of an image; performing feature dimension reduction on the feature data set by adopting a projection matrix method, constructing a projection matrix maintaining a geometric structure, mapping the projection matrix to a low-dimensional potential subspace, and further constructing a model for the subspace by utilizing a sparse self-representation method; constraint is applied to sparse self-representation in the low-dimensional potential subspace, and joint optimization is carried out through an augmented Lagrange multiplier method ALM and an alternating direction minimization ADM strategy to be used for efficient clustering of data in the low-dimensional potential subspace. According to the method, the projection matrix maintaining the geometric structure is constructed, the high-dimensional PRNU features are mapped to the low-dimensional potential subspace, the local neighborhood relation and the global distribution structure are reserved in the dimension reduction process, the calculation cost is reduced, and the clustering robustness and performance are effectively improved.

The invention discloses a semantic-driven intelligent agent capability discovery method and device oriented to the intelligent agent Internet, and the method comprises the steps: firstly generating an intelligent agent structured portrait which covers the three-dimensional information of skills, roles and states; semantic coding is carried out on the portrait, and the portrait is converted into a high-dimensional semantic vector; then, the vector is partitioned, a codebook is generated in each subspace in a clustering mode, and the codebook comprises representative vectors and serial numbers of the representative vectors; and quantizing the sub-vectors into numbers based on a codebook, and splicing the numbers to form discrete identification codes of the agent index. And when the index is updated, incremental maintenance is executed according to the distance between the new agent sub-vector and the vector in the codebook. Meanwhile, a generative retrieval model is trained, task query is directly mapped into discrete identification codes, historical and new task samples are mixed for continuous learning in training, and stability constraints are introduced. According to the method, through semantic portraits, a quantitative index mechanism and memory enhancement continuous learning, an end-to-end retrieval and rapid updating capability discovery scheme is constructed.

The invention discloses an underwater hyperspectral clustering method based on a bipartite graph, and relates to the technical field of underwater image processing. The method comprises the following steps: acquiring spectral images of a to-be-processed image in different spectral intervals, taking the spectral image of each spectral interval as a view, and constructing hyperspectral data based on the acquired views; constructing a hyperspectral clustering model based on subspace clustering learning, and processing the hyperspectral image based on the hyperspectral clustering model; the hyperspectral clustering model comprises an adaptive dynamic anchor point selection module, an anchor point centroid learning module and a bipartite graph decomposition module; and iterating the hyperspectral clustering model based on an alternative update variable strategy until an iteration stop condition is met, and outputting a clustering result based on the clustering indication index matrix. According to the method, a model based on subspace clustering learning is constructed, uniform self-adaptive anchor points are dynamically learned in all pixel points, and the optimal cluster with accurate pixel division can be directly obtained from bipartite graph decomposition.

The invention belongs to the technical field of analysis of load fluctuation behaviors, and particularly relates to an electrical load analysis method, system and device based on subspace clustering-CatBoost and a storage medium. Electrical load data are collected and subjected to standardization processing; clustering the standardized data through a subspace clustering algorithm to generate a clustering label; the clustering labels are classified through a CatBoost classifier, and the fluctuation behavior of the electrical load is analyzed; by performing standardization processing on the electric load data, the inconsistency of data scales is eliminated, the stability of data input is improved, different types of load modes are automatically identified through a subspace clustering technology, the classification of the load data is more accurate, meanwhile, the calculation efficiency of data processing is improved, and through a CatBoost classifier, the calculation efficiency of data processing is improved. Different types of load modes are accurately classified, the intelligent level of electricity consumption prediction is improved, and the interpretability of the model is enhanced.

The invention relates to a consistent semantic enhanced single-step parameter-free multi-view clustering method, which comprises the following steps of: 1) automatically learning anchor points, and avoiding the problem that the quality of the anchor points is reduced due to the randomness of an anchor point selection strategy; (2) spectral clustering is converted into decoupling decomposition of a representation matrix, and single-step data processing is completed without depending on a subsequent clustering method; and 3) the method does not contain any parameter, so that the problem that the clustering quality depends on the parameter is eliminated. According to the method provided by the invention, the accuracy which can be achieved on a Dermatology data set is greatly improved on the same data set compared with the accuracy which can be achieved on the same data set through traditional anchor point-based multi-view subspace clustering. According to the method provided by the invention, the problem that parameters such as multi-view clustering anchor point selection are difficult to adjust can be effectively solved, the problem of optimizing flow splitting is solved, and the method does not need to depend on a subsequent clustering method, so that the multi-view data clustering precision is improved.

The invention relates to a deep embedded subspace clustering method based on double depolarization contrast learning. The method comprises the following steps: selecting a public image data set as a sample set and constructing a clustering model; the method comprises the following steps: firstly, pre-training an auto-encoder, then enhancing data of a data set, inputting the data into the encoder to obtain embedded representation of the data set, and clustering the data set by using a k-means method to obtain an initialized subspace base; secondly, training a clustering model, to be more specific, enhancing data of a data set, inputting the data into two encoders to obtain embedded representations of the enhanced data, performing debiased positive sampling and debiased negative sampling on the embedded representations, and reconstructing the embedded representations of the two enhanced data through a decoder; the distribution probability of each embedded representation in each subspace is calculated by utilizing a subspace base, so that the clustering category of each sample is determined. According to the method, the autoencoder is trained by optimizing the positive and negative sample structures and the sample reconstruction distance, and meanwhile, the subspace base is iteratively optimized, so that a high-quality clustering model is obtained.

The invention provides a multi-view subspace clustering method (TNMVSC) based on a non-convex tensor nuclear norm. According to the method, the low-rank representation theory (LRR), the graph regularization technology and the tensor decomposition theory are combined, and the global structure and the local attribute of each view can be captured at the same time. Besides, the TNMVSC decomposes each low-rank representation sparse matrix into three factor matrixes through a matrix three-factor decomposition theory so as to realize alignment of the representation matrixes, thereby ensuring that the generated core matrix can effectively retain key information. Meanwhile, a non-convex low-rank tensor nuclear norm is adopted to capture high-order correlation among a plurality of core matrixes. In the aspect of algorithm optimization, on the basis of the established optimization model, an alternating direction multiplier method (ADMM) is adopted to carry out optimization solution on the representation matrix of each view. And then, performing angle correction on the fused representation matrix to obtain a similarity matrix among the samples, and clustering the similarity matrix by using a spectral clustering method to realize clustering of the samples. In order to verify the effectiveness of the TNMVSC, tests are performed on reference data sets in multiple fields, including computer vision, bioinformatics, social media analysis and the like. A large number of simulation experiment results show that the TNMVSC is excellent in clustering precision and robustness, and the performance of multi-view clustering can be effectively improved.

The invention discloses a multi-view subspace clustering method, device and equipment and a storage medium, and relates to the technical field of multi-view subspace clustering, and the method comprises the following steps: mapping multi-view data through a plurality of feature mapping matrixes to obtain a plurality of embedded anchor maps; decomposing the plurality of embedded anchor maps to obtain a plurality of internal anchor maps; stacking the plurality of internal anchor maps to obtain a three-order tensor with tensor nuclear norm constraint, rotating the three-order tensor, and applying the tensor nuclear norm constraint based on tensor singular value decomposition to the rotated three-order tensor to obtain a target function; solving the objective function to obtain a plurality of inherent anchor maps after iterative optimization; fusing the plurality of inherent anchor maps after iterative optimization to obtain a similar anchor map; tensor singular value decomposition is applied to the similar anchor images, and a clustering result is obtained. According to the method, anchor points with higher quality are captured, meanwhile, the high-order relation between different views is explored and improved, and the accuracy of a data class cluster division result is improved.

This invention discloses a multi-view subspace clustering method based on diversity graph fusion. The method comprises the following steps: Step 1: Acquire multi-view data and perform preprocessing; Step 2: By adding regularization terms for multi-view consistency and diversity, introduce self-expression learning to explore the intrinsic structure of the multi-view data, and introduce low-rank and sparse constraints on the consistency expression matrix, thereby obtaining a highly reliable and robust similarity matrix; Step 3: Using an induced self-weighting approach, fuse the view similarity matrices obtained in Step 2 to form a final consistent similarity matrix, which serves as the input of a spectral clustering algorithm and outputs the clustering results. This method can improve clustering performance and achieve optimal clustering results.

This invention belongs to the field of radar signal sorting technology, specifically relating to a radar signal sorting method, program, device, and storage medium based on improved sparse subspace clustering. This invention constructs a complex network of radar pulse sequences based on the self-representation properties between radar pulse data points in the same subspace. It utilizes a generalized orthogonal matching pursuit algorithm to reduce the number of connections between pulses from different radars in the complex network, thereby reducing computational complexity. Borrowing the idea of ​​graph segmentation, it uses Laplace spectral decomposition to mine the geometric correlations of radar pulse signals from the same source, achieving efficient sorting of non-ideal radar signals under complex electromagnetic environments. Simultaneously, this invention designs an algorithm based on an energy correction severance threshold, which can effectively solve for the accurate calculation of the number of cluster centers. This invention can effectively solve the problem of mis-sorting caused by aliasing or missing parameters of complex radar operating conditions.

The application discloses a kind of concrete aging state evaluation methods based on intelligent sensor, it is related to intelligent sensor technical field, including: acquisition multi-source sensing data, pre-processing generation window characteristic vector sequence;Carry out environmental mirror difference entropy calibration, form continuous sample set;Sparse subspace clustering is executed, and state cluster label and reconstruction error are obtained;Implement time series gradient entropy driven event mapping, event sequence is generated in joint gradient entropy;Improved hawks process model is constructed, and continuous surface and trigger coordinate parameter are output;Matching stage risk weight matrix is generated state evaluation result;Form aging file, generate early warning information.The application realizes the stable quantitative evaluation and early warning of concrete aging state by the collaborative processing of sparse subspace clustering and improved hawks process intensity modeling.

The invention provides a malicious software package detection method based on script deep semantics, which comprises the following steps: acquiring an original software package compressed package set, extracting an installation script file set and generating a code complexity feature data set; based on the installation script file set, extracting a sensitive behavior context set under the abstract syntax tree; based on the sensitive behavior context set, generating a fusion feature vector set under a CodeBERT pre-training model; obtaining a complete feature vector set based on the code complexity feature data set and the fusion feature vector set; performing subspace clustering based on the complete feature vector set to generate a data space with an abnormal identifier; calculating an abnormal score data set based on the data space with the abnormal identifier, and generating a data space with the abnormal identifier and an outlier identifier under a preset outlier threshold value; and performing malicious detection on the compressed package set of the original software package based on the data space with the abnormal identifier and the outlier identifier. According to the technical scheme provided by the invention, the accuracy of detecting the malicious software package is improved.

The invention provides a point cloud normal estimation method and device based on local subspace clustering, electronic equipment and a storage medium, and the method comprises the steps: introducing a feature point weight evaluation mechanism based on covariance features and geometric structure indexes into a local neighborhood, and screening out feature points with high confidence; carrying out multi-subclass division and plane fitting on the local point cloud by adopting low-rank subspace clustering with priori weight constraint in the neighborhood of the feature points, so as to re-estimate the normal direction of the feature points on the subclass level; meanwhile, noise point filtering and a normal diffusion strategy based on adjacent non-noise points are combined, so that the method can more accurately describe a local geometric structure in a complex scene with noise, multi-structure aliasing and non-uniform sampling, the precision and stability of point cloud normal estimation are remarkably improved, and the method is suitable for popularization and application. And sharp features and edge details of the model can be better maintained.

The invention belongs to the technical field of machine learning and data mining, and particularly relates to a large-scale high-dimensional image data set-oriented compressible subspace clustering method, which comprises the following steps of: firstly, designing a dictionary representation learning model based on a partitioning mechanism to select part of samples to construct a small-scale dictionary to replace the whole original data; a bipartite graph construction method is ingeniously introduced by utilizing the thought of joint clustering, the problem that the bipartite graph cannot be directly constructed due to the fact that a coefficient matrix is not a square matrix is solved, and the relevance between a dictionary sample and a new input data sample can be fully considered. Under the Laplacian matrix rank constraint of the combination graph, the method can directly learn to obtain an optimal structured bipartite graph, can directly obtain a final clustering result, and does not need any post-processing process. In addition, an efficient optimization algorithm based on alternate iteration is further designed in combination with an augmented Lagrangian multiplier method to solve the model.

The present invention discloses a fast deep multi-view subspace clustering method based on dynamic anchors, including: obtaining multi-view sample data, inputting the multi-view sample data into a pre-trained network model for reconstruction and processing, initially obtaining anchor point features of each view, and updating the parameters of the autoencoder and the dynamic anchor network; based on the pre-trained network model, constructing a fine-tuning network, learning the sparse representation relationship matrix of each view, using a self-weight network to adaptively assign weights to the sparse representation relationship matrix of each view, obtaining the weighted sparse representation relationship matrix of each view, fusing the weighted sparse representation relationship matrix of each view, obtaining a multi-view consensus sparse representation relationship matrix, and obtaining a final clustering result through a fast spectral clustering method. The present invention reduces the quadratic and cubic space-time complexities of the existing method with respect to the number of samples to linear overhead, realizing efficient clustering of multi-view data.

The invention discloses a multi-geophysical field joint inversion method based on physical subspace decoupling, and belongs to the field of geophysical exploration. According to the invention, the problem of poor inversion effect caused by difficult characterization of multivariate and nonlinear physical relations of different types of rocks under complex geological conditions in the existing structure coupling joint inversion method is solved; according to the method, K subspace clustering is utilized to dynamically divide underground model units into different low-rank subspaces according to physical property parameter characteristics in each inversion iteration, and then a low-rank constraint based on a Schatten-p norm is applied to a physical property parameter matrix in each subspace; the clustering-coupling process and the inversion iteration are alternately carried out, so that the division of the physical property subspace and the model updating in each subspace are mutually promoted, the self-adaptive optimization is realized, and the clustering-coupling process and the inversion iteration are carried out at the same time; and finally, high-precision and high-resolution reconstruction of multi-type rock physical property distribution in the complex underground structure is realized.

The invention provides a multi-view subspace clustering cancer subtype identification method based on self-reinforcement learning, and the method comprises the steps: firstly, extracting the potential feature representation of each view from multi-omics data through a potential feature learning module; then, clustering similar samples by using a self-expression learning module, and introducing initial graph information as a supervision signal to construct a self-expression coefficient matrix; secondly, inputting the matrix into a view image fusion unit, and fusing multi-view information to generate a consensus image; in addition, in order to further suppress noise interference in multi-omics data, a self-strengthening back propagation unit is introduced, a confidence matrix is generated by optimizing a self-expression coefficient, fusion loss back propagation is guided, the quality of the self-expression coefficient is iteratively improved, and a consensus graph is optimized; and finally, based on the optimized consensus graph, realizing cancer subtype identification by applying a spectral clustering algorithm. According to the method, the self-reinforcement learning strategy is introduced, the interference of noise on sample relation capture is effectively relieved, and the clustering performance is remarkably improved.

The application relates to a social network text target topic detection method based on sparse subspace clustering, and belongs to the technical field of computer data processing of natural language processing and social network data mining.The method writes text identification into target platform text content and interaction event records, carries out word segmentation, denoising and vectorization processing, reduces short text noise and event mismatch interference on subsequent analysis, constructs a social relationship graph, calculates edge confidence weight to form a graph regular constraint parameter, reduces the weight of a low-confidence screen edge in the constraint, suppresses relationship noise caused by organized manipulation from the source, solves sparse representation coefficients, constructs a similarity matrix, performs spectral clustering, enhances the separability of samples with similar semantics but different propagation modes, calculates semantic cohesion and propagation deviation, determines a target topic cluster by using a double threshold, extracts a key text set and a propagation evidence set, forms a reviewable target topic detection result, and improves target topic detection rate and reduces false positives.

The invention discloses a multi-view subspace clustering method for enhancing tensor Schatten P-norm, and relates to the field of multi-view data subspace clustering, the method comprises the following steps: firstly, modeling noise data, and simultaneously considering Laplace noise (l1-norm represents loss) and Gaussian noise (l2, 1-norm represents loss) in the data; noise pollution in the data is effectively removed, and a cleaner tensor structure is obtained. Then, a weighted tensor Schatten P-norm is applied to the denoised tensor, which not only approaches the original structure feature of the tensor, but also fully considers the contribution of different singular values to the clustering performance. According to the method, a tensor denoising model and a weighted tensor Schatten P-norm are integrated into a unified framework, so that effective analysis of complex data is realized, different singular values are endowed with different weights, and remarkable structural features in the data are highlighted.

The invention relates to the field of ideological and political course classification teaching, in particular to a subspace clustering method and device for ideological and political course teaching and a storage medium. The method comprises the following steps: extracting numerical feature vectors of ideological and political course data, performing normalization processing to obtain an input matrix, performing clustering analysis on the input matrix through a subspace clustering method to obtain a low-rank coefficient representation matrix, constructing an affinity graph, and based on the affinity graph, obtaining a clustering result by applying a spectral clustering algorithm. Dividing the original data set of the ideological and political courses into different clusters; according to the method, a non-convex truncation norm is used to constrain low-rank representation and a feature matrix at the same time, a truncation singular value square operator with soft rank constraint is used to enable a low-rank representation matrix to meet the size of a target rank, and a globally optimal solution is obtained in a closed form. The complexity and diversity of ideological and political course data are effectively dealt with, the accuracy and practicability of ideological and political course classification are improved, and implementation of ideological and political course personalized teaching and ideological and political course optimization is supported.

The invention discloses a kernel norm-based robust flexible multi-view subspace clustering method and system. The method comprises the steps of mapping original image data to a high-dimensional regeneration kernel Hilbert space through a kernel technique; learning a consistency kernel through a self-weighted multi-kernel learning strategy; introducing low-rank constraint into self-expression learning of the Hilbert space to obtain a similar matrix; and calculating in the similar matrix according to a k-means algorithm to obtain an image clustering result. According to the method, original image data are mapped to a high-dimensional regeneration kernel Hilbert space by adopting a kernel technology, so that nonlinear data are effectively processed, weights are automatically distributed for different kernels, the importance of different views is effectively considered, the influence of noise is considered, low-rank constraints are added in a similarity matrix, and the image quality is improved. Therefore, an optimal similarity matrix with a clear block diagonal structure is constructed, and a better image clustering result can be obtained.

The present application discloses a motor bearing fault diagnosis method and device based on weighted sparse subspace clustering. The motor bearing fault diagnosis method based on weighted sparse subspace clustering includes: obtaining a sample set, where the sample set includes vibration signals of bearings under different fault types; obtaining the vibration acceleration signal of the bearing to be diagnosed; performing feature extraction on the vibration acceleration signal of the bearing to be diagnosed, so as to obtain a feature vector to be diagnosed; obtaining a weighting term according to the vibration acceleration signal of the bearing to be diagnosed and the sample set; performing clustering through the weighted sparse subspace clustering method according to the weighting term, the feature vector to be diagnosed and the sample set, and obtaining a clustering result. In the algorithm of the present application, the similarity degree between data is introduced as the weighting term, which is beneficial to accurately diagnosing bearing faults.

The invention discloses a mining area illegal expansion remote sensing change monitoring method based on sparse subspace clustering, and relates to the technical field of mining areas, and the method comprises the following steps: obtaining mining area image data, carrying out the preprocessing of the mining area image data, obtaining consistent image data, and extracting an image feature tensor; constructing a self-expression model with sparse constraint based on the image feature tensor, performing subspace division, and extracting a target change region set according to a subspace division result; performing spatial superposition comparison on the target change area set and pre-obtained mining area legal boundary data to generate a newly added change area set, and performing index extraction to obtain a change area quantitative index set; and judging an illegal expansion risk level of the newly added change area set, and generating a change monitoring report according to the illegal expansion risk level and the consistent image data. According to the invention, through illegal expansion of the risk marking layer, a supervision department is supported to carry out rapid positioning, grading management and graph and certificate output.

The invention relates to the field of machine learning and pattern recognition, and particularly discloses a variance-covariance subspace distance-based multi-kernel subspace clustering method, which comprises the following steps of: S1, acquiring a high-dimensional data matrix, calculating a variance-covariance matrix and second-order statistical information thereof, defining a preliminary VCSD kernel value and obtaining a VCSD kernel matrix through exponential transformation and normalization, constructing a multi-core pool of r base cores; s2, constructing an optimized objective function based on non-negative matrix factorization, and solving to obtain a local affinity feature map of a block diagonal structure under constraint conditions; s3, capturing high-order feature association through a matrix method, and converting tensor optimization into matrix operation; s4, solving the target function by adopting an alternating optimization strategy, reducing the calculation complexity, and generating a target affinity graph; and S5, clustering the target affinity spectrum, and outputting a result. According to the technical scheme provided by the invention, the problems of missing statistical characteristics, complex calculation and nuclear noise interference of the existing method are solved, and the clustering precision and efficiency are effectively improved.

The invention relates to the technical field of satellite data processing, and discloses a method and system for recovering satellite state data through historical enhancement, and the method comprises the steps: calculating the similarity between current satellite data and multiple segments of historical data, and selecting a plurality of segments of most similar historical data as matched historical data; designing a tensor fusion filling scheme, determining the similarity between the current satellite data and the historical data, respectively distributing different weights to the current satellite data and the matched neighbor data, fusing the current satellite data and the neighbor data with the distributed weights, training a data completion model, and obtaining a data fusion filling model; outputting complete satellite data through the data completion model; according to the method, the precision of the completion model is effectively improved, the distance between sparse data and historical data is efficiently calculated by adopting a subspace clustering quantization method and utilizing a distance table, the calculation efficiency is improved while the high retrieval precision is kept, and the retrieval performance is remarkably improved.

The present invention provides a search engine-based reference material and standard product retrieval and sorting, which relates to the technical field of data processing, and includes: obtaining search keywords input by a user, constructing a reference material and standard product knowledge base, mapping to a low-dimensional semantic space, calculating the semantic relevance of the search keywords and reference material knowledge, constructing a heterogeneous information network, generating a comprehensive relevance measure, performing multi-granularity semantic matching, and obtaining candidate search results; assigning adaptive feature weights to structured features corresponding to the candidate search results, performing independent clustering, obtaining subspace clustering results, performing optimization, obtaining a global optimal clustering result, mining high-order semantic association features and performing cross-category semantic association, and generating high-quality search results; performing low-rank decomposition, obtaining an initial comprehensive score, modeling the sorting problem as a Markov decision process, determining an action space and a state space, constructing a reward function and determining a sorting position, and obtaining an optimal sorting result.

The present invention discloses an object defect detection method, system, device and medium, relating to the field of artificial intelligence; a feature extraction model is used to extract the features of a target image set to obtain an image feature set; a subspace clustering model with group sparse constraints is constructed, and an objective function and constraint conditions are determined; an iterative method is adopted to solve the objective function according to the constraint conditions to obtain the value of the model parameters when the subspace clustering model has the minimum clustering loss; after determining the cluster centers of the subspace clustering model according to the model parameters, the distances between the image feature set and the cluster centers are calculated to obtain the defect detection classification results of each target object in the target image set; by combining the method of machine learning and computer vision of clustering algorithms, the present invention can achieve accurate detection of defects.

The present embodiment provides a subspace clustering method, apparatus, equipment and storage medium, the method comprising: performing feature extraction, self-representation processing, decoding processing and other operations on the original data features through a preset neural network model to obtain a target loss function, wherein the nuclear norm is used to impose a low-rank prior on the preset self-representation coefficient matrix for self-representation processing. The neural network model is trained through the target loss function, the target self-representation coefficient matrix is ​​extracted from the trained convolutional autoencoder network model, and the target similarity matrix is ​​further obtained, and then the target similarity matrix is ​​segmented using a spectral clustering algorithm to obtain a subspace clustering result. The present application realizes the learning of the self-representation coefficient matrix based on the nuclear norm and the convolutional autoencoder network model, thereby obtaining a similarity matrix with higher accuracy, and finally obtaining a subspace clustering result with higher accuracy. The method of the present application can improve the accuracy of subspace clustering.