175 results about "Supervised clustering" patented technology

The invention relates to the technical field of network security, in particular to a network security big data state evaluation method based on pattern recognition, which comprises the following steps of: extracting multi-modal features from a network flow log, a system event log, a host behavior log and threat intelligence data, generating a feature matrix, performing feature dimensionality reduction by adopting an auto-encoding network, and obtaining a network security big data state evaluation result; carrying out attack behavior classification and abnormal mode identification in combination with unsupervised clustering and a graph neural network; constructing an attack transition probability matrix based on a Markov model; forming a time sequence attack chain; predicting an attack development trend; and a dynamic protection instruction is issued to the safety equipment. According to the method, the unknown attack detection capability can be improved, the time sequence attack traceability is enhanced, the security situation assessment is optimized, and the method is suitable for security situation awareness in cloud computing, industrial internet and large-scale network environments.

The invention discloses an e-commerce platform high-risk user identification method based on time series data analysis. The method comprises the following steps: S1, constructing a user behavior data set; s2, organizing the collected original user behavior data into a behavior sequence according to a time sequence based on the collected original user behavior data set; s3, performing data cleaning on the generated user behavior sequence; s4, carrying out clustering analysis on the preprocessed behavior sequence by adopting a time sequence clustering algorithm in Scikit-learning, and identifying a user group with similar behavior characteristics; s5, dividing the user behavior modes into different risk categories based on a clustering analysis result; s6, automatically labeling new potential high-risk users on the basis of labeled high-risk user data by utilizing a semi-supervised clustering algorithm, and expanding a risk identification range; s7, generating a corresponding risk score; and S8, performing collective management and prevention for high-risk user groups. According to the method, the user groups with similar high-risk characteristics can be identified, and the flexibility and effectiveness of the system in coping with different risk scenes are improved.

A computer implemented method detects and classifies anomalies in an imaging dataset of a wafer comprising a plurality of semiconductor structures. The method comprises determining a current detection of a plurality of anomalies in the imaging dataset, and obtaining an unsupervised or semi-supervised clustering of the current detection of the plurality of anomalies. Based on at least one decision criterion at least one cluster of the clustering is selected for presentation and annotation to a user via a user interface. An anomaly classification algorithm is re-trained based on the annotated anomalies. A system for controlling the quality of wafers and a system for controlling the production of wafers are also disclosed.

The invention belongs to the technical field of hyperspectral image unsupervised, and discloses a hyperspectral image self-supervised clustering method based on a local-global double-branch network, which comprises the following steps: (1) constructing a 3D pixel block; (2) acquiring shallow depth features through a ResNet module; (3) injecting shallow layer features into a dual-path network module; (4) calculating the similarity between the local features and the clustering center to obtain local semantic probability distribution, and then obtaining target distribution; (5) the global features are sent to a feedforward neural network to obtain the probability of each pixel, and global semantic probability distribution is obtained; and (6) constructing a network loss function through a dual self-supervision mechanism to guide the updating of the whole network model. According to the method, the local detail features and the global semantic features of the hyperspectral image are mined by introducing the dual-path network architecture, so that the recognition precision of the hyperspectral ground features is effectively improved.

The system and method for processing a digital pathology image using a machine learning model that includes a self-supervised hierarchical Vision Transformer (ViT) configured to perform unsupervised clustering with multiple classification tokens. The method includes receiving a digital pathology image that depicts a tissue slice stained with histological dyes. The digital pathology image may be processed to generate a result comprising multiple predicted classifications of individual patches of the digital pathology image. The result is generated by a machine-learning model using a self-supervised hierarchical Vision Transformer (ViT) that may further comprise a multi-head self-attention module configured to predict a crosspatch relevance metric using an attention mechanism for each individual patch in the digital pathology image thereby assigning the individual patches to a cluster based on the crosspatch relevance metrics.

The invention discloses a data labeling method and system based on user behaviors and attention tracking, and the method comprises the steps: synchronously collecting multi-source behavior signals of a mouse, a keyboard, eye movement and the like of a doctor in real time, combining identity and interface metadata, and carrying out the standardized normalization, abnormality elimination and short time sequence behavior unit division. And extracting individual behavior micro-modes by using unsupervised clustering, and constructing a behavior portrait library. Through multi-modal time sequence modeling and a self-adaptive space-time attention mechanism, behavior characteristics, an interface area and a report text are deeply fused, a multi-level correlation probability is output, and high-precision automatic tagging of content and an image area is realized.

The invention relates to a construction method of an unmanned aerial vehicle remote sensing image classification model. Performing structured representation of the remote sensing image based on hierarchical feature decomposition; the method comprises the following steps: dividing an unmanned aerial vehicle image into image blocks of different levels through a pyramid decomposition algorithm, performing hierarchical combination for features of each level, and forming a multi-scale structured feature vector through a feature weighting mechanism; capturing dynamic features fused with a spatial topological relation: after a graph structure is constructed, performing feature aggregation and spatial relation information propagation on each node by using a graph convolutional network GCN, so that the features of each node comprise own information and the influence of fusing surrounding ground features; for multi-tense image data, a dynamic graph convolutional network D-GCN is introduced, modeling is carried out on spatial relations at different moments, and changes of ground objects along with time and influences of the ground objects on classification are captured; self-supervised clustering learning of ground feature features: for ground features which are difficult to distinguish, a soft clustering method is introduced, so that the ground features belong to a plurality of categories at the same time;

The invention relates to the technical field of data processing, and discloses a flue gas treatment intelligent regulation and control method and system based on artificial intelligence and a medium. The method comprises the following steps: collecting parameter data of a flue gas system through a sensor network and processing to obtain preprocessed data; using unsupervised clustering to identify working condition features and categories; constructing an emission concentration error active switching control model according to the working condition information; different working condition control parameters are optimized by adopting a differential evolution algorithm; optimized parameters are input into a controller, and modes are dynamically switched according to pollutant errors. The control strategy can be automatically switched according to the working condition characteristics of the flue gas treatment system, the complex and changeable industrial production environment can be effectively dealt with, and the system operation cost is optimized while stable and up-to-standard emission of pollutants is guaranteed.

Systems, methods, and other embodiments associated with quasi-supervised clustering for activity pattern characterization and anomalous activity detection are described. In one embodiment, a method generates a first sparse similarity matrix for nearest neighbors of a plurality of data points. The data points each characterize a pattern of activity associated with an account. The method generates a second sparse similarity matrix for random neighbors of the plurality of data points. The method recursively clusters the plurality of data points based on the first sparse similarity matrix. The method quasi-supervises the recursive clustering based on the second sparse similarity matrix to stop the iterative clustering when the data points are split into N clusters. The value of N is not pre-determined. The method detects that the individual data point has changed clusters, indicating anomalous activity. And, the method generates an electronic alert that the anomalous activity is associated with the account.

The invention discloses a self-supervised clustering method for massive single cell data. The method comprises the following steps: 1) downloading single cell RNA sequencing data from a disclosed gene expression database; 2) preprocessing the single-cell RNA-seq data, wherein the preprocessing comprises filtering, standardization and feature selection; (3) carrying out preliminary clustering on the preprocessed single cell RNA sequencing data by utilizing a rapid Louvain algorithm; 4) screening representative cell points from the preliminary clustering result by using an index of clustering representativeness; 5), after representative cell points are screened out from each preliminary cluster, a Transform model is used, and a final clustering result is obtained.According to the method, the rapid Louvain algorithm and the Transform neural network are combined, the efficiency and accuracy of clustering analysis are remarkably improved, and an efficient and accurate solution is provided for analysis of mass single cell data.

The invention belongs to the technical field of civil aviation data analysis, and particularly relates to an aviation trajectory semi-supervised clustering method based on a Transform noise reduction auto-encoder. Comprising the following steps: S100, preprocessing a flight arrival and departure trajectory data set; s200, data set feature extraction; s300, constructing a model and training the model; s400, carrying out a semi-supervised clustering algorithm; and S500, training the model constructed in the step S300 to obtain a new test set, extracting the data features of the entry and departure and runway numbers according to the step S200, inputting the track data into the model, and clustering the output noise reduction track by a semi-supervised clustering algorithm to obtain a clustering result. According to the method, data mining is carried out on the flight path data of the airspace of the terminal area, a flight path clustering model is constructed, the mode category to which the flight path belongs is identified, and then reference is provided for fine safe operation of the airport terminal area.

The invention discloses a radiation source new individual identification method based on signal enhancement double contrast learning, and belongs to the field of software radio. According to the method, a multi-physical domain enhancement operator sequence is adopted, so that the risk that feature distortion is possibly introduced in traditional data enhancement is overcome, and a high-quality feature discrimination basis is provided for subsequent comparative learning by constructing positive and negative sample pairs with strong correlation. A feature decoupling architecture of double comparative learning is adopted, and the category discrimination advantage of supervised comparative learning and the feature decoupling capability of unsupervised comparative learning are organically fused. Dynamic semi-supervised clustering is realized by adopting a dual decision clustering mechanism, firstly, the number of potential categories is determined by utilizing density peak clustering DPC, secondly, a semi-supervised decision module is introduced to realize known category accurate identification and unknown category adaptive discovery of radiation source individuals, and then new radiation source individual identification is realized. The method is suitable for the field of software radio, and high-precision radiation source new individual identification is realized.

The invention relates to the technical field of safety monitoring artificial intelligence, and discloses a safety management method and system based on artificial intelligence and big data. The method comprises the steps that multi-modal real-time monitoring data is collected through a distributed sensor array, and a standardized sequence is generated after preprocessing; the system accesses a historical security event library, adaptively divides risk category clusters by using an unsupervised clustering algorithm, and extracts key risk features and influence domains for each cluster. On the basis, a plurality of special self-adaptive security analysis engines are configured to perform parallel analysis on real-time data to generate a plurality of security analysis feature maps. And carrying out weighted fusion and enhancement on the feature maps through a feature fusion network to obtain a unified security situation feature vector, and finally outputting a security level and a management and control instruction through a decision model. According to the method, autonomous identification, dynamic evaluation and accurate response of complex risks are realized, and the intelligent level of safety management and the decision accuracy are improved.

Systems, methods, and other embodiments associated with quasi-supervised clustering for activity pattern characterization and anomalous activity detection are described. In one embodiment, a method generates a first sparse similarity matrix for nearest neighbors of a plurality of data points. The data points each characterize a pattern of activity associated with an account. The method generates a second sparse similarity matrix for random neighbors of the plurality of data points. The method recursively clusters the plurality of data points based on the first sparse similarity matrix. The method quasi-supervises the recursive clustering based on the second sparse similarity matrix to stop the iterative clustering when the data points are split into N clusters. The value of N is not pre-determined. The method detects that the individual data point has changed clusters, indicating anomalous activity. And, the method generates an electronic alert that the anomalous activity is associated with the account.

A method and device for labeling transient voltage stability samples in a power grid based on semi-supervised learning are provided. The method includes: S1: obtaining a transient voltage time series trajectory V formed for each load bus in the power grid under N transient operating scenarios; S2: preliminarily labeling the stability status of each transient operating scenario with a voltage time series dataset V, and integrating the labeling result Yi into a class label dataset Y; S3: constructing a voltage stability sample set S={(Vi, Yi)|1≤i≤N}, dividing S into sample subsets Su and Sk; S4: labeling samples in Su by using a semi-supervised clustering learning method and a semi-supervised classification learning method to obtain result datasets Yu1 and Yu2 respectively; S5: performing interactive verification on Yu1 and Yu2, and updating Su and Sk; and S6: performing repeated iteration on the S4 and the S5.

The invention discloses an optical module fine-grained health degree assessment method, system and device and a medium, and the method comprises the steps: data collection and preprocessing: carrying out the semi-supervised clustering, and preliminarily dividing the data into fault samples and normal samples; the method comprises the following steps: carrying out fine granularity division on a normal sample through a Transform auto-encoder; performing fine-grained division on the'fault 'samples through respective Gaussian mixture models of the'fault' samples and the completely healthy samples; judging whether the port state triggers a preset event, and if not, directly outputting an evaluation result; and if at least one preset event is triggered, updating the model for classification in each stage, carrying out health degree grade re-evaluation, and outputting an evaluation result. According to the method, a five-level health degree quantitative standard is established, the full life cycle state of the optical module from complete health to irreversible serious fault is covered, and the method can adapt to partial data drift caused by aging of the optical module.

The invention provides a double-current comparison and DHI combined equipment fault detection method and system, and belongs to the technical field of power equipment state monitoring and fault diagnosis. The method comprises a time sequence perception adversarial enhancement module, a generative adversarial network (GAN) data expansion module, a double-flow contrast attention network (D-CAN) feature extraction module and a dynamic health index (DHI) calculation module. The core lies in that a fault sample with time sequence correlation is supplemented through a physically constrained GAN, robust features are extracted by using a ResNet1D and Transform fused double-flow network, and a health state is quantified in combination with unsupervised clustering and mahalanobis distance. Through simulation and experimental verification, zero-delay detection of the early fault of the transformer can be realized, the output health index and the 3D visualization result can provide an accurate basis for operation and maintenance of the transformer, and the diagnosis accuracy, the data utilization rate and the dynamic adaptability of state evaluation are remarkably improved.

Techniques are disclosed herein that encompass image pre-processing and a semi-supervised clustering for optimization and analysis of immune-enriched single-cell proteomics data generated via multiplexed imaging technologies. This is achieved through an image pre-processing pipeline, which converts image data contained in one type of file (e.g., .mcd) into another type of file (e.g., .tiff) and removes artifact signals from the image data using various algorithms to generate improved image data. Thereafter, a semi-supervised clustering pipeline analyzes the improved image data using various techniques, including implementing a supervised algorithm to identify metaclusters such as general immune phenotypes (e.g., CD4−T-cells, Macrophages, Neutrophils, etc.) as well as non-immune phenotypes while implementing an unsupervised algorithm that enables the identification of specific subclusters and a more in-depth cellular status characterization.

The invention discloses an unsupervised silicon wafer surface defect detection method and device based on hierarchical clustering, relates to the field of image processing and computer vision, and utilizes an unsupervised clustering technology to detect silicon wafer surface defects. The method comprises the following steps: extracting multi-scale features from a plurality of normal silicon wafer images through a pre-trained convolutional neural network; integrating the multi-scale features of the plurality of pictures by adopting a hierarchical fusion strategy; constructing a feature memory library by using a clustering algorithm, and generating a clustering center; generating patch features for the test image through a consistent fusion process; through comparative supervision, normal patches are close to a clustering center, and abnormal patches are pushed away; calculating the distance between the patch and the clustering center, and judging abnormity; and mapping the position of the abnormal patch, and positioning scratches, cracks, pollution or recesses. According to the method, through an unsupervised mode, a large amount of defect labeling data is not needed, the data preparation cost is effectively reduced, and the accuracy and robustness of silicon wafer surface defect detection are remarkably improved.

The invention discloses an intelligent calibration method and system for the electric quantity of an instrument, and relates to the technical field of electric energy metering of an electric power system.The intelligent calibration method comprises the steps that an electric meter network diagram with a main meter and sub-meters as nodes is constructed, and robust composite correlation indexes are calculated on the basis of reading increments in a sliding time window; the interference of common background load and instantaneous noise is suppressed by using a threshold clipping and main table purification mechanism, so that the depiction capability of the correlation graph on a physical connection relationship is improved fundamentally, and the phase to which the sub-table belongs can be identified more accurately in a stock area with unknown phase or incomplete topology; a graph neural network model is introduced based on the network graph, electric meter embedding representation with physical meaning is obtained through neighborhood information aggregation, phase grouping is achieved in combination with unsupervised clustering, split-phase energy balance loss is designed in the training stage, and the difference between the reading of a main meter and the sum of the reading of sub-meters in each phase group is used as a constraint. Alignment of the embedding space and the energy conservation relation is pushed in a self-supervised mode.

The invention discloses an academic community discovery and analysis method driven by a citation network, and the method comprises the steps: constructing the citation network, constructing an adjacent matrix A and a node attribute matrix X according to the citation network, and extracting a core sub-network information matrix S based on a k-core algorithm; constructing a dual-channel sparse graph attention auto-encoder, respectively taking (X, A) and (S, A) as input of the two channels to learn low-dimensional embedding of the two channels, and obtaining joint embedding Z through a dynamic weighted fusion strategy; reconstructing adjacent matrix information by adopting an inner product decoder, reconstructing node attribute characteristics and core sub-network information, and calculating reconstruction loss; and inputting the joint embedded Z into a self-supervised clustering module, performing joint optimization on the information embedding and reconstruction module and the self-supervised clustering module, and finally outputting a group division result of the papers / authors. According to the method, complementary fusion of attributes and a core structure is realized, robustness in noise and sparse scenes is enhanced, and discriminability and stability of group division are improved.

The invention discloses an automatic driving test scene generation method and device based on a text collision report, equipment and a medium, and relates to the technical field of automatic driving, and the method comprises the steps: obtaining an unstructured text collision report, converting the unstructured text collision report into a high-dimensional numerical vector through preprocessing and feature extraction, and obtaining a high-dimensional numerical vector; a typical collision mode is identified and classified by using an unsupervised clustering algorithm, a logic scene is further constructed in a parameterized manner, physical feasibility is verified through a kinematics verification module, risk levels are divided, and a risk scene library is formed. And finally, generating an automatic driving test case set by adopting a stratified sampling strategy. By using the collision text data, real, diversified and physically feasible test scenes are generated, and the efficiency and accuracy of automatic driving safety assessment are improved.

The invention provides an edge deletion network community detection method based on dual-channel deep embedding clustering, and the method comprises the steps: firstly proposing three edge deletion strategies which are used for simulating an edge deletion condition possibly occurring in an actual network; then, an edge enhancement algorithm based on random walk is put forward to recover missing edge information under different edge deletion strategies; a clustering model based on double-channel depth embedding is introduced and comprises an information embedding core module and a self-supervised clustering core module. According to the method, the missing edge condition in a real network is simulated by providing three different edge deletion strategies, and the missing edge is effectively recovered in combination with an edge enhancement algorithm based on random walk, so that the robustness and accuracy of community detection in an edge missing scene are greatly improved. Network core node information and high-order neighbor information are respectively subjected to embedded representation, node features with higher distinction degree are obtained through a dynamic weighted fusion strategy, and more sufficient information support is provided for subsequent community division.

The invention provides a drawing projection marriage matching test system based on a visual converter and a method thereof. According to the system, drawing behaviors, emotion feedback and interaction data of a user are collected, multi-modal feature extraction is carried out through a visual converter, and drawing styles, psychological features and emotion preferences of the user are deeply analyzed through a self-supervised clustering algorithm. A unique emotion feedback loop is embedded into a system, and micro expressions, gestures and emotion changes shown by a user in the drawing process are captured in real time, so that more accurate personalized marriage matching recommendation is realized. The system generates a comprehensive love and marriage matching report, provides customized love and marriage suggestions in combination with potential characters, behavior modes and emotional requirements of users, and is matched with emotional management and mental health support tools. The innovative system improves the scientificity, interactivity and deep emotion association of love and marriage matching, has high application value, and can be widely applied to the fields of love and marriage platforms, emotion consultation and intelligent mental health.

The invention discloses a geothermal fluid source discrimination method based on geochemical index and machine learning fusion, and relates to the technical field of geothermal resource intelligent exploration, and the method comprises the steps: firstly collecting and calculating multi-dimensional geochemical indexes, and constructing a standardized feature sequence; establishing an end member feature library by using unsupervised clustering; identifying the source type of the fluid through a discrimination model fused with an attention mechanism; aiming at the mixed source fluid, constructing an optimization model embedded with physical and chemical constraints, and quantitatively inverting the contribution proportion of each end member; and finally, on the basis of the time sequence proportion data obtained by inversion, predicting a future evolution trend by adopting a time convolutional network-long and short-term memory network model fused with an attention mechanism. According to the geothermal fluid source discrimination method based on geochemical index and machine learning fusion provided by the invention, the whole-process intelligent analysis of the geothermal fluid source from qualitative identification to quantitative prediction is realized.

The invention relates to the technical field of computers, in particular to a low-temperature economizer digital twinborn body construction method, and aims to solve the problems that an existing model statically solidifies, multi-source data fusion is difficult, and degradation recognition lags. The method comprises the following steps: constructing a multi-physical field reference model covering fluid, heat transfer and corrosion mechanisms; collecting and carrying out time-space alignment on temperature, pressure, flue gas components and ash deposition data, and combining wavelet and median filtering to carry out de-noising; introducing extended Kalman filtering to correct model state variables on line, and realizing dynamic updating of parameters; a CNN-GRU hybrid network is embedded to extract time sequence degradation features, and unsupervised clustering is combined to identify working condition states; and when the detection is abnormal, triggering local high-fidelity CFD re-simulation, and completing closed-loop optimization and state synchronous mapping. Through the mechanism, the internal physical field error lt of the equipment is realized; 5% high-precision dynamic mapping, more than 95% of anomaly detection rate and rapid deployment of a new unit within 72 hours are realized, the prediction accuracy and preventive maintenance capability are remarkably improved, and the service life of equipment is prolonged by 15%-20%.

The invention discloses a channel environment measured data traveling wave fault distance measurement method, which belongs to the technical field of channel environment traveling wave fault distance measurement, and comprises the following steps of: 1, screening traveling wave fault recording data; step 2, calibrating a traveling wave head; step 3, carrying out power transmission line traveling wave fault distance measurement; and step 4, optimizing the traveling wave distance measurement device. According to the method, fractal dimensions and transient energy are used as feature vectors, a small number of known category historical samples are introduced as prompts, partial supervision clustering is formed, the fault membership degree of weak faults is improved, and blindness of clustering number and initial value selection is avoided. The continuity of a single fault event action and the tightness in time are considered, the time neighborhood constraint is added in the clustering process, and the detection capability of small-fault-angle faults and high-resistance faults far away from a clustering center recognition sample is enhanced. Interference waves are eliminated according to the length and slope characteristics, and reliable calibration of the wave head is achieved.

The invention discloses an autonomous system organization mapping method and device based on a semi-supervised clustering algorithm, and the method comprises the steps: extracting the original data information of an autonomous system from a preset dimension, and carrying out the preprocessing of the original data information; integrating the preprocessed original data information to generate an autonomous system basic feature matrix and an autonomous system key feature constraint matrix; according to an SC-MP I algorithm, performing multi-round iterative clustering on the autonomous system basic feature matrix and the autonomous system key feature constraint matrix, and determining the number of clustered class clusters in a binary approximation mode to obtain a plurality of class clusters; and according to the characteristic mode of the autonomous system in each class cluster, determining the organization name and code of the class cluster, and obtaining a final autonomous system-organization mapping relationship. It is ensured that the data dimension is wide enough; the model can be easily deployed to a stable production environment, and excessive manual intervention on data is not needed; and the flexibility of the clustering mapping process is improved.

The invention relates to an intelligent color box printing color analysis control method, and provides a reinforcement learning decision closed-loop control method integrating distributed color detection, space-time multi-modal data processing, feature extraction and self-supervised clustering and multi-scale reward modeling. Multi-variable normalization and synchronization are completed through multi-point real-time collection of color, process and environment data, color change meta-state features are extracted based on a self-encoding neural network, and a meta-state distribution model related to space and working conditions is established through self-supervised clustering. And the color risk is evaluated in combination with a multi-scale reward function, so that color parameter optimal adaptive adjustment and cooperative control driven by reinforcement learning are realized. In an actual control link, the system can dynamically optimize a meta-state aggregation and strategy network structure based on a feedback signal, a stable and self-adaptive color consistency closed-loop control system is constructed, and the color consistency and process intelligence level of a printed product under a complex working condition are effectively improved.

The invention discloses a text deep clustering-based community discovery method, which comprises the following steps of: 1, training a text by utilizing a training model to generate a word vector, and converting the preprocessed text into a word matrix; 2, extracting global context information of the text and local information between continuous words, splicing to obtain text feature vectors, calculating cosine similarity of the text feature vectors, and constructing a text network; step 3, performing community discovery based on clustering to obtain an initial community center node and a community discovery result; and step 4, constructing a deep self-supervised clustering loss function, updating the text network and the community center by minimizing the loss function, and obtaining an updated text feature vector and a community discovery result. The problems that an existing community discovery method singly considers global and local features of a text and model parameters cannot be optimized according to a community division result in the community discovery process can be solved.