74 results about "Small data sets" patented technology

The invention relates to a sensitive data identification and desensitization method and system based on an AI large model, and belongs to the technical field of data security processing and artificial intelligence. The method comprises the following steps: receiving a multi-source data processing request, establishing a sensitive data identification framework, analyzing elements of sensitive data, and extracting a sensitive data feature vector through a natural language processing algorithm; classifying the multi-source data, dynamically adjusting an identification threshold value of the sensitive data identification framework through a sensitive data feature vector, and generating a self-adaptive sensitive data identification strategy; establishing a terminal desensitization verification model, performing desensitization verification on the identified sensitive data, performing mask replacement or generalization processing according to privacy requirements, and outputting a desensitization retry processing scheme; the sensitive data is subjected to block desensitization, a large data set is divided into small data sets for independent desensitization, replaceable entities in information are removed according to data attributes, and the desensitized data is further subjected to secondary desensitization.

The invention relates to the technical field of radar signal processing, in particular to a ground penetrating radar reinforcing steel bar clutter suppression method based on physical constraints, and mainly solves the problem that a data set is difficult to obtain in an existing ground penetrating radar reinforcing steel bar clutter removal method based on deep learning. The method is an improved method based on the CUT network, the CUT network structure and a comparative learning mechanism determine that the requirement of the network for the data size of a data set is low, a waveform smoothness constraint is added on this basis, the clutter removal effect and generalization ability are improved by introducing physical prior, the physical constraint serves as a regularization item, and the regularization efficiency is improved. The problem that a CUT network is prone to model collapse under a small data set is solved. Finally, the improved model is compared with other models through different evaluation indexes, and the result shows that the improved model has more advantages.

The first token prediction of a large language model is bottlenecked by compute and second token predictions onwards are bottlenecked by memory bandwidth. Inferences can be made more efficient through activation sparsity. An activation tensor is pruned using an importance threshold value. The mode of the activation tensor is centered in a lossless manner using an estimated mode value to improve activation sparsity further. Pruning and mode-centering mechanisms can be inserted into a neural network strategically and post-training to implement sparsification. A two-stage greedy grid search algorithm is implemented to determine the calibrated importance threshold values of various pruners and the estimated mode values using a small dataset. A modified neural network with pruning and lossless mode-centering can be deployed onto hardware.

The invention provides a reverse design method of a composite function metasurface based on transfer learning, relates to the field of metamaterial design and optimization, and solves the problem of low efficiency of a traditional design method of a metasurface with a complex structure. Comprising the following steps: generating a data set containing various metasurface structure parameters and corresponding spectral responses; building a deep prediction neural network taking the unit structure parameters as input and the spectral response as output; the trained network is combined with a genetic particle swarm optimization algorithm, metasurface phase distribution, transmissivity and other requirements determined in advance are used as objective functions, and optimal parameters are searched through multiple iterations; and migrating universal features learned on the large-scale data set by the source domain network model through a parameter fine tuning method, applying the universal features to a new small data set target domain, repeating the previous process until all parameters are determined, and completing the design. According to the method, the complex metasurface design efficiency can be improved, the neural network performance is ensured, meanwhile, the requirement of the deep learning method for the data collective quantity is reduced, the robustness of the model is improved, and a low-cost practical scheme is provided for metasurface inverse design.

The present invention relates to the field of organic polymers. Specifically disclosed is a method for predicting the properties of grafted polypropylene for a high-voltage power cable. In the method of the present invention, a property prediction model for grafted polypropylene is constructed on the basis of a multi-layer perceptron model and a random forest model. The model solves the defect that a conventional machine learning model mainly relies on a large amount of data and is not suitable for a small data set. Experimental data is used to quantitatively establish a relationship between a grafted structure and mechanical properties, the amount of data required is small, the data is easy to obtain, and the prediction accuracy is high, thereby providing a reliable grafted group for a grafted polypropylene material used for a high-voltage power cable.

The application belongs to the field of data enhancement application, and particularly relates to a GAN model improvement method for small data set scenarios. The multi-discriminator fusion module is composed of multiple discriminators and a discriminator fusion submodule. Each discriminator is a network composed of multiple convolutional layers and fully connected layers, which receives a batch of pictures in a downsampling manner and converts the pictures into a tensor. The tensor has four dimensions, namely the number of batches of images batch, the number of channels c, the width w of each image and the height h of each image. The multi-discriminator fusion submodule is responsible for coordinating multiple discriminators, receiving the return values of the multiple discriminators, weighting the return values and then transmitting the weighted return values to the generator for back propagation. In order to avoid the occurrence of sample penetration, the weight of the discriminator responsible for enhancement is reduced. In this way, the probability of overfitting of the discriminator can be reduced, and the possibility of sample penetration can be reduced.

An improved multi-target regression algorithm based on target specific features is combined with a non-dominated sorting genetic algorithm to construct a double-layer structure multi-target algorithm, so that the problems of data distortion possibly caused by data enhancement, poor feature engineering generalization ability, negative migration caused by large inter-domain difference and the like are avoided, and the method is suitable for large-scale popularization and application. By capturing the correlation among a plurality of output targets, a high-precision model for a high-dimensional small data set is constructed, and the machine learning assisted hot-rolled ultrahigh-toughness steel component and process design with low data set cost is realized.

The application belongs to the technical field of unmanned aerial vehicle image detection, and in particular to an unmanned aerial vehicle image detection method based on an improved RT-DETR model. The method comprises the following steps: selecting a public unmanned aerial vehicle image dataset, dividing the training set, constructing a small dataset, training and evaluating the improved RT-DETR model on the small dataset, and then implementing the model on the complete dataset to obtain the final data. The application enhances the information of small targets in the fusion process without increasing the number of model parameters. While introducing S2, a new fusion module mainly composed of DSConv is used to replace the fusion module of the benchmark model, which balances the parameter quantity of the model while improving the small target detection accuracy. Compared with the previous algorithm, the improved model has higher accuracy and smaller parameter quantity, and can provide protection for accurate detection of unmanned aerial vehicle images.

The application relates to the technical field of archive construction, and provides a text attribute relation extraction method based on a small data set pre-training model, a pre-training model based on an ERNIE structure is publicly fine-tuned to a relation extraction model in a business background through a public pre-training model UIE which can be started by fine-tuning only a small amount of labeled data, and on the basis of the model, Before_pipeline data processing before prediction and After_pipiline data processing after prediction are constructed in combination with business text features to improve result reliability, the problems of lacking a labeled data set, UIE violent sentence splitting leading to errors, and low extraction result accuracy of the UIE are solved, and the problem that an existing traditional relation extraction model depends on a large amount of high-quality data sets manually labeled is solved.

Methods and systems for ultrasound imaging are provided. In one example, a method includes receiving, with a wireless handheld probe assembly, ultrasound signals of a region of interest; generating, within the wireless handheld probe assembly, a plurality of received digital signals based on the received ultrasound signals; generating each of a larger data set and a smaller data set from the plurality of received digital signals; transmitting the smaller data set from the wireless handheld probe assembly to a hub via a lower bandwidth wireless connection; transmitting the larger data set from the wireless handheld probe assembly to the hub via a higher bandwidth wireless connection; generating, at the hub, a low resolution image from the smaller data set and a high resolution image from the larger data set; and transmitting the low resolution image from the hub to a first display and the high resolution image from the hub to an electronic device.

The application discloses a kind of based on mechanism and data driven's sugarcane squeezing process process index prediction method, comprising the following steps: (1) multi-sensor field data acquisition and preprocessing, establish original data set;(2) establish sugarcane material elastic-plastic constitutive model;(3) establish sugarcane squeezing process porous medium control equation;(4) sugarcane squeezing process fluid-solid coupling model and its simulation, and establish new data set;(5) the establishment of physical guide neural network model;(6) the training details of physical guide neural network, obtain based on mechanism and data driven's sugarcane squeezing process process index prediction model.The method of the application is mainly used to predict the key process indexes such as sugarcane juice precipitation amount and sugarcane juice precipitation speed in the sugarcane squeezing production process.The cost of data acquisition in the production process is greatly reduced.High-dimensional system is characterized by small data set;The consistency of network can be greatly improved, and the generalization ability is enhanced.

The application provides a semi-supervised surface defect instance segmentation method based on pseudo label enhancement, comprising the following steps: obtaining a surface defect image dataset; training an initial instance segmentation network with a small amount of labeled defect image data; inputting the labeled defect image data into a student instance segmentation network for training; inputting unlabeled defect image data into a teacher instance segmentation network and the student instance segmentation network to obtain the mask and features of each defect instance; then inputting the mask and features into a cross-supervised contrast learning module and a general distribution fusion module; constructing a contrast learning loss function, obtaining fused pseudo labels, training the student instance segmentation network with the pseudo labels, and updating the parameters of the teacher instance segmentation network using the parameters of the student instance segmentation network. The application can cope with the challenges of small data set size and fuzzy defect boundary, estimate complex class feature distribution through the general distribution fusion module, and improve the quality of the mask pseudo labels from the teacher model.

The application discloses a kind of multi-dimensional dynamic model under the evaluation method of soil salinization to cultivated land productivity, it is related to the field technical field of agricultural resources and environment, and the multi-source data set of fusing remote sensing, ground monitoring, meteorological and farmer management data is first constructed in the scheme, and the minimum data set of soil physical, chemical, biological properties is screened after fusion preprocessing;Soil quality index (SQI) and cultivated land productivity index (CPI) are calculated again;Subsequently, the multi-dimensional dynamic evaluation model containing key salinization index screening, effect quantification, mechanism model and crop sub-model is constructed;Finally, the productivity influence dynamic update is realized in combination with real-time data of Internet of Things, and the optimal improvement technology mode is simulated and screened by structural equation model simulation management scene. The method is comprehensive in data coverage, accurate in evaluation, dynamic in model and targeted, can reveal the influence mechanism of salinization on cultivated land productivity, provide practical decision support, and help efficient and sustainable use of saline-alkali soil.

Systems and methods are provided for an electronic match engine of an exchange that distributes a minimum data set to an external market data generation (MDG) processor. The electronic match engine derives a minimum data set from data already known to the electronic match engine. The MDG processor, which may be outside of the electronic match engine, may extract the minimum data set and uses it to generate market data. The electronic match engine may append the minimum data set to an order entry message sent to the MDG processor.

The application provides a process parameter screening method and device for an anti-penetration high-entropy alloy, and relates to the field of armored materials.The method comprises the following steps: obtaining performance indexes based on theoretical and target shooting experiment analysis of an alloy material target plate under normal penetration, and taking heat treatment process parameters of solid solution and aging as characteristic factors to design an orthogonal experiment to obtain a data set; performing standardization processing on characteristic values in the data set, and performing transformation on the characteristics; performing redundancy judgment and selection on time characteristics and temperature characteristics before and after the transformation; establishing a machine learning model, and performing generalization performance evaluation on the model through a root mean square error and a determination coefficient; and establishing a multi-objective optimization method based on a Pareto front and a multi-dimensional joint probability distribution function to realize multi-objective fusion and screening of process parameters.The data set in the application is completely obtained from experiments, which reduces generalization errors generated when a machine learning model is built for a small data set, and can accurately screen required process parameters in a short time.

The application provides an efficient unbalanced PSI based on a Bloom filter and a hash, wherein in a PSI calculation process, a Bloom filter is used to perform one round of screening on a large data set to reduce the overall complexity; a step of sending Bloom filter parameters by a large data set party to a small data set party; a step of calculating, by the small data set party, an index set of bit positions required to be set to 1 in the Bloom filter according to the data set of the small data set party, and sending the index set to the large data set party; a step of initializing the Bloom filter by the large data set party according to the Bloom filter parameters and the index set of the small data set party; a step of screening, by the large data set party, using the Bloom filter to obtain an element set that may be in a private intersection in the data set of the large data set party, wherein the screened set is used as a new private data set for subsequent calculation of the private intersection; and compared with an original PSI based on a hash and a semi-honest third party, the efficiency is greatly improved in an unbalanced scenario.

The application provides a method for predicting the band gap of porphyrin MOF based on deep transfer learning. First, a small data set of porphyrin MOF is constructed, and the Average SOAP Kernel is used to expand the data set of porphyrin MOF. Then, the GCN is pre-trained by using the QMOF database to obtain a pre-training model. The pre-training model is fine-tuned by using a transfer learning method to obtain a porphyrin MOF band gap prediction model. Finally, the band gap is predicted by using the porphyrin MOF band gap prediction model. The application can predict the band gap of porphyrin MOF with high accuracy. The application combines deep learning and transfer learning to predict the band gap of porphyrin MOF, and only depends on crystallographic information files instead of manually made geometric descriptors and chemical descriptors. The band gap of a given porphyrin MOF is efficiently and accurately predicted, the prediction accuracy and efficiency are improved, and the time and economic cost of measuring the band gap of porphyrin MOF are reduced.

The invention discloses a network planning method and device based on a hybrid reinforcement learning strategy, and the method comprises the steps: firstly, providing a basic network structure and an adjustment strategy through building a data link network topological structure model and a strategy support library; and then, an optimization strategy model is constructed based on multi-relation constraints, and efficient network planning decision optimization is performed through a hybrid reinforcement learning strategy so as to cope with a complex network environment. And then, generating an initial network planning structure through small sample data input, and finally, performing security review through the trusted network planning decision model to obtain a target network planning structure conforming to the standard. According to the method, a mixed strategy reinforcement learning method is adopted, high-dimensional target optimization and agile and efficient decision making under the condition of dynamic environment change are achieved, an intelligent planning technology based on a small data set and empirical knowledge is introduced, and the credibility and safety of network planning decision making are enhanced.

The application relates to the technical field of battery management, and specifically discloses a kind of energy storage battery power estimation method and device based on multiple incremental features, electronic equipment and medium, comprising the following steps: obtaining test data of battery under multiple ambient temperatures and operating conditions;Incremental feature extraction is carried out from the obtained test data, and a joint feature matrix is formed with the original test data;The joint feature matrix is pretreated to obtain pretreated data;The pretreated data is processed using a sliding window technique to obtain power evaluation data;The power evaluation data is input into a pre-constructed energy storage battery power estimation model based on multiple incremental features for iterative training to obtain a pre-trained SOC estimation model;The pre-trained SOC estimation model is migrated to a small data set battery to obtain the power of the small data set battery.The application not only improves the accuracy of SOC estimation, but also has strong adaptability and stability, providing strong support for the practical application of the battery management system.

The application discloses a patent technology function extraction method for technical literature, analyzes the features of patent technology terms and function terms in the field of high-end equipment, and improves the accuracy of technology term and function term extraction in Chinese patents. The application constructs a deep learning model for technology term and function term extraction, combines the sentence pattern rules of patents, constructs the heuristic features of technology terms, positions the function sentences by constructing a function term feature dictionary, accelerates the training speed of the model, and improves the extraction accuracy. In order to reduce the cost of manual sample labeling and avoid the model overfitting phenomenon caused by too small data set, a self-training algorithm is used to realize weak supervision learning of the model. The technology terms most similar to the theme of the patent text are selected from the word clustering, the cosine similarity is used to combine the similar semantic terms, and a technology function matrix is constructed.

The invention discloses an automatic attack countermeasure method based on a multi-target memetic algorithm, which comprises the following steps: sampling after sorting loss values from a complete data set, and selecting a small data set for rough evaluation; initializing a population, generating an initial population containing a plurality of random attack sequences, and evaluating the robust precision and time consumption of each individual by using a small data set; generating a progeny population through crossover and mutation operations, and evaluating individuals in the progeny population; performing non-dominated sorting and crowding degree distance calculation on individuals of the parent population and the descendant population, and selecting a certain number of excellent individuals to form a current population; optimal individuals in the current population are randomly selected, local search is carried out by adjusting a loss function, an iteration step length or a restart point, and iteration is continuously carried out until a preset termination condition is not met; and carrying out local search on a better individual by utilizing the complete data set, adjusting a loss function, an iterative step length or a restart point, and outputting an optimal adversarial attack combination mode.

The application discloses a self-adaptive building day-ahead load prediction method based on transfer learning and relates to the technical field of buildings and environmental protection.The application comprises the following steps: S1, data acquisition and processing, wherein original data sets are divided into small data sets of target buildings and large data sets of basic building groups, and missing values of all original data sets are filled; S2, use mode clustering; S3, source domain data screening, historical daily load curves of use modes of load target buildings are screened, and a data transfer training set and a model transfer training set are respectively constructed; S4, day-ahead load prediction model construction; and S5, self-adaptive model optimization, wherein model parameters are continuously adjusted by using Bayesian optimization, and self-adaptive load prediction of target buildings is realized.The application realizes load prediction of target buildings by combining historical data of data-sufficient building groups with the data transfer and model transfer methods of transfer learning.

The invention relates to a method for automatically generating CT (Computed Tomography) volume data to MRI (Magnetic Resonance Imaging) volume data based on a multi-dimensional diffusion framework, which is characterized in that an improved multi-dimensional diffusion generation framework is constructed, and the improved multi-dimensional diffusion generation framework comprises a two-dimensional extensible diffusion model and a three-dimensional extensible potential diffusion model which are modified by a diffusion module and are arranged in sequence; constructing multi-dimensional training data by using original paired CT volume data and MRI volume data, and inputting the multi-dimensional training data into the improved multi-dimensional diffusion generation architecture for training; and inputting CT (Computed Tomography) volume data into the trained improved multi-dimensional diffusion generation architecture to generate MRI (Magnetic Resonance Imaging) volume data. According to the method, input data of any shape can be accommodated, a multi-dimensional diffusion generation framework is provided, a detailed result is realized by using a two-dimensional extensible diffusion model and a three-dimensional extensible potential diffusion model, and the method exceeds the most advanced mode conversion technology based on deep learning; additional training data is allowed to be generated from a small data set, and the amount of data is enriched.

The present invention discloses a two-stage sparsification method for rapidly sparsifying a large language model based on a small sample, comprising the following steps: S1, mask initialization: initializing the mask of the language model using a plurality of mask initialization methods based on a plurality of importance indices for measuring parameters; S2, mask reordering: reordering the masks based on a small sample calibration set and a straight-through estimator; S3, mask adjustment: based on the small sample calibration set and the reordered mask matrix of step S2, using the original model as a teacher model and the sparse model as a student model, implementing a block-level self-distillation process of the large language model before and after sparsification in the time dimension to obtain a sparse model; while achieving comparable performance, this method consumes less computing power, uses a smaller data set, occupies less GPU memory, and does not affect the inference speed of the sparse model, and has significant practical value and application prospects.

The present application provides a federated forgetting method based on malicious terminal intervention training, which belongs to the technical field of privacy computing and federated learning. The present application eliminates the influence of malicious clients on the global model through federated forgetting, and subtracts the parameter update of the malicious client from the final global model parameter generated by federated learning, so as to continue training with a low-quality model that is theoretically infeasible to save the time of retraining, so that the server can delete the influence of the malicious client more quickly when performing the forgetting operation without seeking the willingness of the client whose contribution is deleted; and a comparison mechanism for judging the effect of the forgetting model of the last round and the effect of the forgetting model of the current round is set to analyze the forgetting effect, so that the forgetting operation is ended in advance to inhibit the influence caused by excessive forgetting of the forgetting model; secondly, a small data set is used to train the last forgetting model to restore the deviation of the model caused by the training process, thereby effectively improving the accuracy of the final forgetting model.

The invention provides a visible light panchromatic and infrared image fusion method and system based on deep learning. The method comprises the following steps: respectively obtaining the characteristics of visible light panchromatic and infrared images by using a pseudo twin network; fusing the visible light panchromatic and infrared image features; and carrying out dimension reduction reconstruction on the feature image to obtain a final fusion image. According to the method, the residual network is added into the pseudo-twin network, and the features of two adjacent layers are spliced in dimension, so that the problem of gradient disappearance during forward transmission of the features on a small data set is avoided; visible light panchromatic and infrared feature images are used as a whole to serve as input of a next module, and the limitation of artificially designing fusion rules can be avoided; the learning ability of the network is improved by establishing a multi-loss function model meeting the requirements of an unsupervised learning fusion task.

This invention relates to a remote sensing extraction method for pear orchard areas based on the Re-UNet model, which overcomes the shortcomings of inaccurate classification results and low efficiency in pear orchard area extraction from remote sensing images compared with existing technologies. The invention includes the following steps: acquiring a remote sensing image dataset; constructing the Re-UNet pear orchard area extraction model; training the Re-UNet pear orchard area extraction model; acquiring and preprocessing the remote sensing images of the pear orchard areas to be segmented; and obtaining the remote sensing extraction results for the pear orchard areas. Based on the UNet semantic segmentation model, this invention solves the overfitting problem that easily occurs in small datasets. It also incorporates spatial and channel attention mechanisms and a residual module, further enhancing the feature transfer and cumulative integration characteristics of pear orchard areas in high-resolution remote sensing images, effectively reducing the "salt and pepper" phenomenon and misclassification, and improving the overall segmentation accuracy.

The invention discloses a multi-model fusion motor imagery intention recognition system which comprises a training fusion strategy module and an application fusion strategy module. The training fusion strategy module obtains an optimal model-parameter group and the output weight of each model-parameter pair in the optimal model-parameter group according to the offline training data set, wherein the model-parameter pair is a combination of a feature extraction method, a classifier and corresponding parameters; the application fusion strategy module is used for training each model-parameter pair in the optimized model-parameter group by utilizing an offline clinical data set based on the acquired offline clinical data set of the user and real-time clinical online data, inputting the clinical online data into the trained optimized model-parameter group, and outputting the clinical online data into the optimized model-parameter group; and based on the output weight of each model-parameter pair, carrying out weighted fusion on the output of all the model-parameter pairs to obtain a motor imagery intention recognition result. According to the method, the motor imagery intention recognition precision can be improved without complicated hyper-parameter optimization under the conditions of a small data set and finite time.

The application belongs to the technical field of single cell sequencing, and particularly relates to a single cell TCR sequencing calculation method. The application firstly sets a sliding rectangle in a plane rectangular coordinate system, then performs difference analysis on multi-group data in the rectangle, and finally integrates the difference analysis results. A sliding window with a suitable width is created according to different data sets, and the narrower the sliding window, the more accurate the analysis. The proportions of different cells in each subgroup are subjected to difference analysis, and the difference results are recorded. The sliding window continues to slide and continuously records, and finally a difference analysis matrix is generated. The difference analysis by the sliding window method can effectively avoid the bias caused by the number of cells, so that more accurate analysis results are obtained. The application has very good analysis effect on small data sets.

The invention discloses an insurance marketing activity multi-label classification method based on an ECC (Error Correction Code) algorithm, and the method can effectively model the dependency relationship between labels, improve the classification accuracy under a small data set, reduce the model overfitting risk, reduce the label dimension through rule pre-classification, and improve the feature generalization ability and the stability of a recommendation system.