664results about How to "Strong generalization" patented technology

The invention relates to the technical field of medical data processing, in particular to an advanced nasopharynx cancer treatment effect prediction system based on deep learning. According to the invention, a data acquisition module acquires multi-modal medical images and clinical data of a patient from a hospital information system; the data preprocessing module is used for cleaning, registering, normalizing and standardizing the data; the feature extraction module extracts image features by using a convolutional neural network and fuses the image features with clinical features to form a comprehensive feature vector; the model construction and training module constructs a deep learning model, and is adjusted and optimized through training; the treatment prediction module inputs the comprehensive feature vector to obtain a treatment effect prediction result; and the result evaluation and feedback module compares prediction and actual results, calculates indexes and collects and feeds back the improved model. The system comprehensively integrates data, accurately mines features, constructs an effective model, provides personalized prediction to assist clinical decision, continuously improves the model performance, and has important significance in improving the prediction accuracy of the treatment effect of advanced nasopharynx cancer, promoting personalized medical treatment and the like.

The invention relates to the technical field of environmental sound event detection, in particular to an environmental sound event detection method based on multi-modal data fusion, which utilizes a microphone array, a camera and a sensor to synchronously acquire audio, video and environmental data to form a multi-modal data source. And the collected multi-modal data is preprocessed, so that the data quality and the accuracy of subsequent analysis are improved. And key features are extracted from the processed data, and the features are fused to form multi-modal features. And a deep learning model is constructed, and multi-modal features are used for training, so that a model capable of identifying an environmental sound event can be obtained. And performing optimization such as pruning, quantification and knowledge distillation on the trained model, and deploying the model on edge equipment to realize real-time preliminary feature extraction and event detection. The problems that an environment sound event detection method in the prior art is poor in robustness, low in real-time performance and limited in generalization ability are solved.

The invention relates to a multisource pavement disease identification method based on a language and an image large model, and the method comprises the following steps: collecting related standards of road pavement disease detection, building a knowledge corpus, and reading a road disease image through a large language model LLM; when it is determined by the LLM that the crack crosses the whole image, cutting is carried out in the direction perpendicular to the crack by passing through the center point of the image, and the image is cut into two images; afterwards, all images and corresponding text prompts are input, target detection is carried out on pavement diseases by using a visual language large model VLM, and a prediction box is output; and performing fine adjustment on the SAM by using the image marked with the disease mask to obtain the RoadSAM adapted to the road disease segmentation task. According to the method, a large language model, a visual language large model and a segmentation large model are combined, corresponding fine adjustment is carried out on the segmentation large model SAM, dependence on a large amount of annotation data is reduced, and meanwhile unified segmentation of multi-source and multi-mode pavement disease images is achieved.

The invention provides a river section water quality prediction method based on a combined machine learning model, and the method comprises the steps: building river monitoring section long-time sequence water quality big data, standardizing and supplementing water quality index data, preparing a water quality index data set and a water quality evaluation data set, and selecting and processing water quality big data features. Developing a water quality prediction model based on a combined machine learning algorithm, evaluating the performance of the water quality prediction model, and testing the generalization ability of the water quality prediction model. The correlation of river section historical data in a time sequence is fully utilized, the model not only improves the accuracy and real-time performance of water quality prediction, but also enhances the response capability to sudden water quality events, the monitoring and early warning cost is remarkably reduced, the efficiency and scientificity of water environment management are improved, and the method is suitable for popularization and application. The self-learning capability and the high-frequency data processing capability of the system ensure timely updating and long-term effectiveness of a prediction result, and provide solid technical support for water pollution prevention and control and water quality management.

The invention relates to the technical field of software reliability guarantee, in particular to a code vulnerability detection method based on efficient parameter fine tuning of a code pre-training model, which comprises the following steps: preprocessing source codes of a vulnerability data set; s2, converting the preprocessed source code, and extracting a token sequence and a code attribute graph path corresponding to the source code; s3, constructing a code vulnerability detection model, wherein the code vulnerability detection model comprises a semantic feature extractor, a structural feature extractor, a feature fusion layer and a classifier; s4, jointly training a code vulnerability detection model by using the token sequence corresponding to the extracted source code and the code attribute graph path; and S5, through the optimized code vulnerability detection model, outputting a result that whether the source code to be detected has vulnerabilities or not. According to the method, the model is finely adjusted through efficient parameter fine adjustment, training resources are saved, good performance is obtained, and the code vulnerability detection efficiency and accuracy are improved by fusing and learning semantic and structural features of source codes at the same time.

The invention provides a lottery store violation abnormity identification method and system, and relates to the technical field of image identification, and the method comprises the steps: collecting multi-source heterogeneous image data in real time through an edge computing device; preprocessing the collected multi-source heterogeneous image data to form a stereo image data set containing space, depth and thermodynamic information; constructing a multi-task deep learning model, wherein the multi-task deep learning model comprises a face recognition module, a behavior analysis module and an environment detection module; inputting the three-dimensional image data set into the trained multi-task deep learning model, and executing face recognition, behavior analysis and environment detection tasks at the same time; based on an execution result of the multi-task deep learning model, generating a violation score by applying a fuzzy comprehensive evaluation method, and weighting by considering historical violation records; and generating a lottery store site investigation plan based on the violation scoring result.

The invention discloses an anomaly detection and repair method for non-Gaussian multi-mode sensor data, and belongs to the technical field of unmanned system multi-source autonomous navigation. The method solves the problems that an existing method is poor in performance when facing non-Gaussian data, anomaly detection is difficult to carry out on high-dimensional unstructured data, the robustness of a static anomaly detection model is insufficient, and redundancy compensation is lacked when abnormal data of a sensor is repaired. The method specifically comprises the following steps: step 1, acquiring IMU, GNSS, point cloud and image data of an unmanned system, and then carrying out preprocessing and time synchronization on the acquired data; step 2, respectively extracting features of IMU, GNSS, point cloud and image data at each acquisition moment after time synchronization; step 3, fusing the data features at each acquisition moment; and 4, performing anomaly detection on the data at each acquisition moment according to a fusion result, and further performing data restoration if the data is detected to be abnormal. The method can be applied to data anomaly detection and repair.

The invention discloses a marketing data link abnormity diagnosis method, system and device and a storage medium. Relates to the technical field of power system data management. At present, the problems of multi-source data format difference, lack of real-time anomaly detection mechanism and low data management efficiency exist in anomaly diagnosis. The method comprises the steps that multi-source heterogeneous data of a marketing data link is collected in real time, and data cleaning, missing value filling and abnormal value processing are completed; extracting statistical characteristics of the link data, performing clustering analysis on the data, and generating a self-supervision label; constructing a support vector machine classification model, and detecting normal and abnormal data in the link; optimizing the anomaly detection model through a deep neural network; and realizing interpretability analysis and decision support of anomaly diagnosis by using a multi-classification result generated by the LSTM model and a self-supervised learning strategy. According to the invention, data link abnormity diagnosis is efficiently and intelligently realized, and data quality management, abnormity detection precision and service decision support capability are improved.

The invention discloses a rainy day target detection method based on paired image data generation and knowledge distillation, and belongs to the technical field of automatic driving target detection. The method comprises the following steps: acquiring a real image of a road scene and a semantic segmentation mask; inputting the semantic segmentation mask and the weather condition text information into a trained improved diffusion model to obtain a road scene generation image under a rainy day condition; introducing a CoT module and an ODConv module into the target detection model for optimization, and then improving the target detection model by adopting a dynamic weighted aggregation loss function to obtain an optimized target detection model; performing knowledge distillation to obtain a rainy day target detection enhancement model; and performing target detection by using the rainy day target detection enhancement model. The method provided by the invention is excellent in performance in a rainy day scene, has good generalization ability, can provide higher-precision target detection in various complex scenes, and effectively improves the detection accuracy and stability.

The invention discloses a novel CAD model feature recognition method based on a graph and a point cloud neural network, and the method comprises the steps: creating a data set containing heterogeneous machining features, converting a B-Rep model into a surface adjacent graph with surface-edge attributes, a surface UV grid, and a surface point cloud input neural network model; the neural network model integrates a graph encoder, a point encoder and a multi-task head, and the graph encoder and the point encoder learn classifiable embedded vectors from shape primitives of different scales of points, edges and surfaces respectively; and the multi-task head converts the embedded vector into corresponding semantic segmentation, instance segmentation and bottom surface segmentation results through three meta-learners. According to the method, insights can be extracted from different representation forms of a CAD model, semantic segmentation, instance segmentation and bottom surface segmentation tasks of heterogeneous processing features are processed in parallel, and experiments prove that Gamp is a Gamp; the PCNet has effectiveness and superiority in non-heterogeneous and heterogeneous processing feature recognition, and has application potential in real industrial scenes.

The invention discloses a short text clustering method and device based on a large language model, and belongs to the technical field of natural language processing, and the method comprises the steps: converting a short text into vector representation, and carrying out the preliminary clustering of the vector representation through a clustering algorithm, and obtaining an initial category; for all the corresponding short texts in each initial category, generating an abstract by using a large language model, generating a plurality of theme titles for summarizing the core meanings of the texts according to the abstract, respectively taking the theme titles as sub-titles, inputting the plurality of sub-titles into the large language model, and carrying out theme combination to obtain a parent title, constructing the parent title and the child titles into a hierarchical theme system; the large language model allocates to-be-classified short texts to one or more categories according to the hierarchical topic system to complete short text clustering; and performing quality evaluation and error pattern analysis on a short text clustering result to optimize a clustering algorithm and a large language model. According to the method, the theme of the short text can be automatically found and classified, and the classification accuracy and efficiency are improved.

The invention discloses a large-model-driven Web task automatic execution method and system, and relates to the field of artificial intelligence. According to the method, Web page content is perceived and analyzed through a large model, a state-action graph is constructed and maintained, a task sequence library is stored and managed, and semantic matching and automatic task execution are carried out. According to the method, the problems of high manual intervention requirement and low task processing efficiency in traditional Web task automation are solved, and the technical effects of remarkably improving the automation level, enhancing the intelligent decision-making capability, perfecting a knowledge memory mechanism and efficiently executing are achieved.

The invention relates to the technical field of subgrade settlement prediction, in particular to a subgrade settlement prediction method based on deep learning, electronic equipment and a medium, the subgrade settlement prediction method provided by the invention adopts a subgrade settlement prediction model for prediction, and the model adopts the combination of a convolutional neural network and a bidirectional long-short term memory network. Spatio-temporal features are effectively captured, and the prediction precision and stability of the model are improved; by introducing a channel attention enhancement mechanism module, dynamic weighting is performed on different feature channels, the selection ability of the model to key features is improved, and the sensitivity of the model to subtle changes in data and the capture ability of the model to global features are enhanced; the model is trained through the genetic algorithm, the efficiency is improved, and the genetic algorithm can automatically find an optimal combination in a complex hyper-parameter space, so that the model can show good generalization ability on different roadbed settlement data sets, and the prediction precision of the model is improved.

The invention relates to the technical field of chemical process fault diagnosis, in particular to a chemical process fault diagnosis method based on multi-scale spatial-temporal feature fusion, which comprises the following steps: regularly acquiring fault data in a chemical production process, preprocessing the fault data and inputting the preprocessed fault data into an MTS-F model; the MTS-F network comprises a time sequence feature extraction module and a spatial feature extraction module which are parallel, and the time sequence feature extraction module comprises a long short-term memory network, a channel attention mechanism module, a one-dimensional convolution layer and a global maximum pooling layer which are connected in sequence; the spatial feature extraction module comprises a GS-CNN module, a maximum pooling layer, two groups of spatial separable convolution layers and maximum pooling layers, and a global maximum pooling layer which are connected in sequence; and after the outputs of the two parallel networks are spliced, a prediction result of fault classification is output through a discarding layer and a full connection layer. According to the method, the time sequence and space features are extracted simultaneously in a parallel mode, the fault diagnosis capability is improved, the model parameter quantity and the calculation quantity are reduced, and good generalization performance is achieved.

The invention provides a multi-parameter GNSS-R soil humidity inversion method based on a random forest model. The multi-parameter GNSS-R soil humidity inversion method comprises the following steps: satellite-borne GNSS-R data quality control; satellite-borne GNSS-R data observation quantity extraction and surface reflectance calculation are carried out; eliminating satellite-borne GNSS-R observation data influenced by the open water area; carrying out surface reflectance gridding; preprocessing the SMAP data; constructing an inversion model; and carrying out spaceborne GNSS-R soil humidity inversion and precision evaluation. The method has the beneficial effects that by utilizing the performance advantages of processing a non-linear relationship, good generalization performance, strong robustness and the like of the random forest, after observation data influenced by a water body is accurately eliminated, an inversion model is constructed by combining multi-parameter auxiliary data of a soil humidity active and passive remote sensing satellite; the space-time coverage, the inversion precision and the reliability of soil humidity data are remarkably improved, and powerful data support is provided for the fields of climate forecast, agricultural production, hydrological modeling and the like.

The invention relates to a product surface defect small sample detection method based on comparative learning. The method comprises the following steps: collecting a defect image sample and carrying out image preprocessing; based on an improved image language comparison pre-training model CLIP, establishing a surface defect detection model; selecting a defect image sample to carry out finetune training on the model; and inputting a defect image to be detected into the trained surface defect detection model, and outputting a defect detection result. According to the method, the strong generalization ability of the image language comparison pre-training model CLIP is fully utilized, high detection accuracy can be achieved only through a small number of defect samples, and the problems that the number of the defect samples is small, and large-scale collection is difficult are solved. According to the method disclosed by the invention, the Q-K-VAttention in the Vision-Transform feature extraction network is improved into the V-VAttention, so that the fine-grained feature extraction capability is enhanced.

The invention discloses a target voice extraction method based on multi-reference clue fusion, and the method comprises the steps: enabling a to-be-extracted multi-channel mixed voice to pass through a feature extraction module, calculating the complex spectrum and spatial features of the mixed voice, and obtaining a mixed voice feature map; enabling the azimuth information and the reference voice of the target speaker to pass through an embedded feature extraction module, and combining the direction features and the voiceprint features to obtain embedded features of the target speaker; inputting the mixed speech feature map and the embedded features of the target speaker into a main extraction network based on a dual-path Fastform neural network to obtain high-dimensional feature representation of the speech to be extracted; the high-dimensional feature representation is converted into a plurality of masks for estimating the separated voice through a mask estimation module, and the masks act on a plurality of spectrums of the mixed voice signals; estimating and extracting a time domain waveform of the speech through inverse short-time Fourier transform reconstruction; and the extracted voice of the target speaker is input into the comparison learning module, so that the correct extracted voice can be output when the mixed voice is absent by the target speaker.

The invention provides a two-stage registration method and device suitable for a full-length image of a lower limb, a terminal and a medium. The method comprises the steps of obtaining a to-be-registered CT image and a label of the lower limb, and a target X-ray image and a label of the lower limb; inputting the to-be-registered CT image and the label, and the target X-ray film image and the label into an affine transformation network to generate an affine transformation matrix for performing rough registration on the to-be-registered CT image; and inputting the CT image and the label output in the coarse registration stage, and the target X-ray film image and the label into a deformation field registration network to generate a deformation field for fine registration based on the coarse registration result of the CT image to be registered. According to the method, the deep learning technology is utilized, a brand-new 2D / 3D medical image registration algorithm is constructed, and full-automatic cross-modal registration of lower limb full-length horizontal position CT and standing position X-ray films is achieved for the first time. The research result of the invention can provide effective assistance for preoperative accurate planning, assist the design work of an HTO operation guide plate, and improve the clinical curative effect of early and middle knee OA.

The invention relates to a lithium ion battery health state estimation method based on an MSCNN, and belongs to the technical field of lithium ion batteries. The method comprises the following steps: performing an aging cycle test on a lithium ion battery monomer, and establishing a working condition test experiment database; collecting voltage, current and temperature data of the battery in each aging cycle, respectively calculating average values, standard deviations, skewness and kurtosis of the voltage, the current and the temperature, taking the average values, the standard deviations, the skewness and the kurtosis as health factors, and screening out the health factors which are highly related to the health state of the battery; constructing an SOH estimation model, and training the model by taking the screened health factors as input; and the model is used for predicting the health state of the lithium ion battery after being trained. The method can effectively extract features from data of different scales, has higher SOH estimation precision compared with a traditional CNN, can perform accurate SOH estimation on data sets of different working conditions, and has good generalization ability.

The invention belongs to the technical field of image recognition, and relates to a small sample learning method for task-adaptive association learning, which comprises the following steps of: S1, assisting the original prompt by using a specific prompt word of a task; s2, enhancing the discrimination of the training features by using comparative learning; s3, using diversified image enhancement means to enhance the contrast of the image; according to the method, the generalization ability of the model is enhanced through rich description of a natural language, so that the CLIP can be more efficient even if the CLIP faces the unmet category or complex concept. Therefore, the method not only reduces the dependence on a large-scale annotation data set, but also improves the learning ability of the model under the condition of small samples and even zero samples, enhances the adaptability and generalization ability of the model to a new environment, greatly simplifies the deployment process of the model, and improves the deployment efficiency of the model. The method becomes a powerful tool for solving a series of challenges in image recognition and description tasks.

The invention, which relates to the technical field of image processing, discloses a method for converting a 2D image into a 3D image, comprising the following steps: step 1, obtaining a 2D image; step 2, monocular depth estimation is carried out on the 2D image to obtain a relative depth map; 3, determining the longitudinal depth and the zero parallax surface of the scene according to the content of the 2D image; 4, carrying out value domain transformation on the depth map; 5, calculating the parallax of each pixel of the 2D image according to the transformed depth map, and then rendering an image of a virtual viewpoint according to the parallax; and step 6, repairing the cavity. According to the method for converting the 2D image into the 3D image, the watching experience is improved, the stereoscopic impression is enhanced, the ghosting phenomenon is reduced, and the proper proportion scale is kept; the conversion efficiency and the automation degree are improved, and manual intervention is reduced; meanwhile, the generalization ability is enhanced; in addition, the hole repairing effect is optimized, so that the converted 3D image is more natural and perfect.

The invention discloses a microcirculation blood vessel automatic segmentation and quantification method based on enhanced feature fusion. The method comprises the following steps: step 1, data preparation and preprocessing; 2, designing an enhanced feature fusion network model; step 3, automatically segmenting microcirculation blood vessels; and 4, quantifying the diameter of the blood vessel based on the segmentation result. According to the enhanced feature fusion network model provided by the invention, accurate segmentation of microvessels is realized through fusion and enhancement of multi-scale features. The model shows good generalization ability on diversified clinical image data through effective data enhancement, and can adapt to image data with different illumination conditions, contrast ratios and resolutions. Therefore, the model has high practicability and reliability in practical application. According to the method, the diameter of the blood vessel is quantified through the segmentation result, the diameter of the microvessel can be accurately measured, and powerful support is provided for evaluation of the clinical microcirculation function.

The invention relates to an edge anomaly traffic detection method based on a multi-scale aggregation Transform, and the method comprises the steps: S1, obtaining edge network traffic data, carrying out the data preprocessing and feature extraction, and obtaining a training sample set; s2, constructing an edge anomaly traffic detection model based on multi-scale aggregation Transform, wherein the edge anomaly traffic detection model comprises an input layer, a coding layer, a classification layer and an output layer; the input layer extracts feature information from the flow data; the coding layer adopts a layered structure to carry out multi-scale time sequence representation, a coding module of each layer is composed of a time sequence representation module and a Transform coding module, the time sequence representation module carries out segmentation mapping on feature data, and the Transform coding module encodes a mapping vector by using a multi-head self-attention mechanism; after the output of the coding module is aggregated, a detection result is obtained through the classification layer and is output through the output layer; training the model through the training sample set; and S3, applying the trained model to an edge system to assist the edge system in detecting and filtering abnormal traffic. The method is beneficial for improving the accuracy of abnormal flow detection.

The invention discloses an image forgery detecting and positioning method based on noise auxiliary prompt learning, which introduces a contrast language image pre-training model (CLIP) and prompt learning into the field of image forgery detecting and positioning, and comprises the following steps of: 1, respectively processing an input image and noise thereof by using an image encoder of the CLIP and a visual transformer to obtain double-branch features, the forgery perception capability of the network is enhanced through multi-domain fusion and a memory mechanism; 2, creating a pair of learnable prompts as negative and positive samples to replace discrete prompts, and finely adjusting the prompts according to the features and categories of each image; and 3, completing forgery detection by calculating similarity between prompt and image features, and integrating image and text features through feature linear modulation to obtain a forgery positioning result. According to the invention, the accuracy and generalization of image forgery detection and positioning can be improved, and the false alarm problem can be reduced, so that the performance of image forgery detection and positioning can be improved.

The invention relates to a flexible workshop scheduling method and system. Comprising the following steps: constructing a flexible job shop scheduling mathematical model according to a flexible job shop scheduling problem; the flexible job shop scheduling mathematical model is converted into a Markov decision process, a state space and an action space are set, and a reward function is obtained; constructing a heterogeneous graph network, defining types and numbers of nodes and edges, and defining a directed relationship between the edges; constructing a heterogeneous perception encoder, and carrying out feature embedding on the heterogeneous graph network to obtain an environment state feature vector; constructing an actor-commentator network model, and inputting an environment state feature vector; and the PPO algorithm optimizes the actor-commentator network model according to a reward function, updates parameters, obtains a flexible workshop scheduling model and completes FJSP solution. According to the method, the resource and constraint relationship in the scheduling environment can be accurately reflected, and the FJSP can be more accurately solved; and the scheduling problem of different scales can be solved, the generalization performance is good, and the method is excellent in large-scale application.

The invention discloses a blood vessel segmentation method and device based on large model fine tuning, equipment and a medium, and the method comprises the steps: obtaining a blood vessel image, carrying out the preprocessing of the obtained blood vessel image, and obtaining a training set; performing fine tuning on a pre-trained image segmentation model by adopting the training set, so that the image segmentation model learns related features of blood vessel segmentation; wherein the low-rank adaptive fine tuning technology is adopted to perform fine tuning on the image segmentation model: two parallel low-rank matrixes are inserted into a Transform layer in the image encoder, the input is processed by the two low-rank matrixes in sequence, and the processing result and the original output are added to obtain the final output. According to the method, the defects of a large model in a blood vessel segmentation task are overcome through efficient parameter fine tuning and a multi-scale feature fusion module. Accurate blood vessel segmentation can be realized under limited annotation data, the difficulty and cost of medical image segmentation tasks are greatly reduced, and the method can be widely applied to the technical field of medical image processing.

The invention discloses an item type self-growth and self-evolution and classification method based on a large model. The method comprises the steps that item type catalogs are added to a large language model, and corresponding item classification cues are constructed; inputting the government affair hotline work order text and the item classification prompt word into the fine-tuned large language model, and quoting an item type directory to output an item type corresponding to the government affair hotline work order text; and outputting the item type output by the large language model as a classification result of the to-be-classified government affair hotline work order text. The item type catalog needs to be periodically updated based on the large language model. Prediction of government affair hotline item types is achieved in the mode that the large language model is combined with the item classification prompt words, meanwhile, fine adjustment is conducted on the large language model so that the large language model can be more focused on the field of government affair hotlines, and item type self-growth and self-evolution of the item type catalog are achieved through the large language model. Therefore, the timeliness and accuracy of government affair hotline item type prediction are improved.

The invention discloses a method and device for predicting the geological storage amount of carbon dioxide and a medium, and relates to the technical field of carbon dioxide capture, utilization and burying. The method comprises the steps that whether a to-be-predicted oil and gas reservoir and a control oil and gas reservoir are in the same block or not is judged; if yes, inputting the geological data, the production dynamic data and the injection parameter combination scheme of the to-be-predicted oil and gas reservoir into the optimal control oil and gas reservoir carbon dioxide geological sequestration quantity prediction model to determine the whole-area accumulated carbon dioxide geological sequestration quantity of the to-be-predicted oil and gas reservoir; if not, geological data, production data and an injection parameter value range of the to-be-predicted oil and gas reservoir are obtained, migration training is conducted on the optimal control oil and gas reservoir carbon dioxide geological sequestration quantity prediction model, and a to-be-predicted oil and gas reservoir carbon dioxide geological sequestration quantity prediction model is obtained; and determining the whole-area accumulated carbon dioxide geological sequestration amount of the to-be-predicted oil and gas reservoir. According to the invention, the precision and efficiency of carbon dioxide geological sequestration quantity prediction are improved.

The invention discloses a reinforcement learning and graph neural fusion algorithm for effectively solving flexible workshop scheduling, which comprises the following steps of: formalizing an FJSP into an MDP by setting states, actions, state conversion and rewards, and describing a scheduling process as dynamically allocating ready operation to a compatible and idle machine; in a scheduling process, firstly converting a scheduling state into a heterogeneous graph structure; carrying out two-stage embedding by adopting a heterogeneous graph attention network, and extracting features of jobs and machines from a heterogeneous graph for embedding; the decision network generates action probability distribution by using feature embedding, and samples scheduling actions from the action probability distribution; and training the deep reinforcement learning model by using a near-end strategy optimization algorithm. Compared with the traditional PDR and SOTA DRL methods, the algorithm has the advantages that a high-quality solution is obviously realized on two synthetic data sets with different distributions, and the performance and generalization ability are improved. Good generalization ability is also exhibited because the model trains large-scale or out-of-distribution instances on small-scale instances.

The invention belongs to the technical field of network security, and discloses an Internet of Things attack detection method based on deep learning, which eliminates redundant features through feature selection, enhances inter-class difference and intra-class similarity of selected feature subsets, solves the problem of class imbalance in an attack detection data set by utilizing EQLv2, and improves the attack detection efficiency. And meanwhile, constructing a space-time model to extract space-time information of the sample. According to the method, multi-classification performance evaluation is carried out on an Internet of Things attack detection reference data set NSL-KDD and a CIC-IDS-2017, the advantages of the method are verified from five indexes including the accuracy rate, the precision rate, the recall rate, the F1 score and the false alarm rate, and the effectiveness and feasibility of the method are reflected when the method is compared with a plurality of methods.