89 results about "Early stopping" patented technology

The invention discloses an attention mechanism and LSTM fused industrial bearing vibration time sequence signal fault prediction method and system. The method comprises the following steps: collecting a bearing vibration signal and carrying out filtering, noise reduction and normalization preprocessing; constructing a deep learning model combining the bidirectional BiLSTM and a coordinate attention mechanism to extract bidirectional time sequence features and enhance key fault features; carrying out model training by adopting a multi-target composite loss function and an Adam optimizer, and introducing an early stop mechanism to prevent overfitting; performing fault type identification and degree evaluation on the real-time vibration signal by using the trained model, and performing quantitative analysis by fusing multi-scale spectrum kurtosis features and nonlinear kinetic parameters; and finally, outputting a fault diagnosis report, and triggering multi-stage early warning based on an adaptive threshold. The method can realize high-precision and high-reliability bearing fault prediction and health state evaluation, and is suitable for intelligent operation and maintenance of industrial equipment.

The invention provides a self-adaptive bearing fault diagnosis method based on multi-base wavelet fusion. The objective of the invention is to solve the problems of noise reduction, insufficient feature extraction and low diagnosis precision under noise conditions. A Kaisixi University bearing public data set is used as original data, and Gaussian noise with different SNRs is superposed to simulate various noise intensities. And uniformly carrying out length alignment, down-sampling, equal-length segmentation, division and normalization preprocessing. Then, wavelet bases such as sym4, db4, coif5 and the like are adopted for parallel multi-scale decomposition and reconstruction; and adaptively determining the number of decomposition layers and a threshold strategy according to the noise level, and generating a de-noising branch. And performing weighted fusion on the denoising results of the branches, and performing iterative denoising on the residual error. Signals subjected to noise reduction processing are sent to a double-branch convolution-cycle-attention network, a convolution layer extracts features, an LSTM and a self-attention module capture time sequence changes, and accurate recognition of various bearing faults is achieved. The training adopts a segmented attenuation learning rate and an early stop strategy, and the robustness and generalization ability of different SNR working conditions are improved.

The invention discloses a PINN-based limited space internal explosion damage parameter prediction method and system. The method comprises the steps of completing data acquisition and constructing a time sequence data set; model construction is completed, and the time sequence modeling capability is enhanced; determining a total loss function and adding boundary condition constraints to enable model prediction to accord with physical laws; firstly optimizing data loss, then introducing a physical residual error, and finally activating a boundary speed suppression loss item and adjusting a learning rate; in combination with a learning rate dynamic scheduling and early stop mechanism, the training efficiency and stability are improved through adaptive residual weighted balance data and physical constraint loss; and predicting parameters such as pressure, temperature and speed of each point in the limited space in multiple time frames by using the trained model, calculating impulse based on a pressure time history, evaluating personnel damage, and completing damage zoning in the limited space. And efficient modeling under a complex boundary condition and a limited space environment is completed.

The invention discloses a visibility regression prediction method based on multi-modal transfer learning and time coding, and relates to the technical field of artificial intelligence. The method comprises the following steps: S1, dividing a data set in different periods according to illumination characteristics, and splitting each period into a training set, a verification set and a test set; s2, preprocessing the data set; s3, constructing an initial model containing a pre-training deep learning network, a time coding module and a multi-layer perceptron regression head; s4, extracting image visual features and time feature vectors; s5, fusing the features and inputting the features into a regression head for prediction; s6, carrying out scheduling training by using layered parameter freezing, an AdamW optimizer and a dual learning rate, and combining with a mixed early stop strategy until convergence; and S7, evaluating the test set to determine a final model. According to the method, complementarity of image and time information is mined, high-precision prediction is realized, generalization is good under different illumination conditions, a layering strategy and an optimization mechanism guarantee stable and efficient training, a multilayer perceptron combination technology enhances expression, and overfitting is effectively prevented.

The embodiment of the invention discloses an air conditioner maintenance data classification method and system based on machine learning, and the method comprises the steps: integrating multi-source heterogeneous maintenance records generated in the maintenance process of air conditioner equipment, building a correlation index through a common identification field, and fusing dispersed data into a maintenance data set in a uniform format; performing hierarchical semantic analysis on unstructured texts in the set to generate structured semantic features, and performing time sequence feature extraction on structured data; then constructing a hybrid classification model training framework fusing semantic and time sequence features, and generating a maintenance data classification model through feature space alignment, dynamic weight distribution, hyper-parameter optimization and an early stop strategy; and finally, classifying newly-added maintenance records by applying the model, checking by combining an expert knowledge base, manually rechecking conflict results, and returning corrected data as an incremental sample back to the model to realize continuous optimization.

The invention discloses a multivariable time sequence prediction method and system based on an implicit neural network. The method comprises the following steps: 1) collecting data and preprocessing the data; 2) performing window division on the standardized or normalized multivariable time sequence and determining the length of a to-be-predicted window; 3) performing variable correlation coding on the input window to obtain a variable-level feature vector; 4) the implicit neural network based on time attention predicts target parameters by using the variable features in the step 3), and implicit neural representation of the target sequence is modeled through the parameters; 5) taking the output of the implicit nerve representation and the original input window as the input of the multi-head attention predictor, and obtaining a prediction result through cross-sequence cross attention calculation performed in the implicit space and multi-layer perceptron conversion output dimension; and 6) training and optimizing model parameters, calculating a mean square error of a prediction result and a real result, taking the mean square error as a loss function, carrying out back propagation to optimize trainable parameters of the variable correlation coding module, the implicit neural network module multi-head attention predictor and the multi-layer perceptron, and then repeating the steps 3) to 6) to obtain the multi-head attention predictor. Until the preset number of iterations is reached or the error of the model on the verification set meets the requirement of early stop; and 7) performing prediction by using a model of training convergence, and performing reverse normalization on a prediction result to obtain a final prediction result. The method has good generalization, and meanwhile, the interpretability of the attention mechanism is remarkably improved by generating the hidden space characteristics of the trend component and the season component.

The invention discloses a multi-degree-of-freedom mechanical arm control method based on quantum particle swarm optimization of a migration strategy. The method comprises the steps that system modeling is conducted; mPC to-be-adjusted parameter definition and constraint processing are carried out; adopting, adopting, and introducing penalty terms to construct a composite fitness function; updating particle positions by adopting a basic or enhanced quantum updating mode; triggering conditions are judged in a group diversity measurement mode, and when the conditions are met, a migration strategy and dynamic subgroup division are executed; the global optimal solution and the fitness value are loaded into a real-time MPC controller, and online adjustment is carried out; a prediction equation is constructed, tracking errors and energy consumption optimization are converted into a standard quadratic programming problem, and the solving precision is dynamically adjusted in combination with Cholesky pre-decomposition, a structured sparse solver and a warm-start and early stop strategy; and a closed-loop adaptive control system is constructed. According to the method, the precision, robustness and real-time performance of trajectory tracking control of the mechanical arm can be effectively improved.

The invention provides a denoising method for a magnetic resonance spectrum based on Transform, and relates to the technical field of signal processing and artificial intelligence, and the method comprises the steps: firstly obtaining a noisy free induction decay signal (FID) and a corresponding label signal, constructing a training data set, converting the signal into a frequency domain signal through fast Fourier transform, and carrying out the amplitude normalization; a real part and an imaginary part of a frequency domain signal are respectively projected to a high-dimensional feature space, and are respectively sent to a multi-head self-attention mechanism-based multi-layer Transform encoder through two channels, so as to capture the global correlation of a long sequence. And the coded output is separated and decoded, a real part and an imaginary part are recovered respectively, and a denoised complex frequency domain signal is obtained through combination. In the training process, a complex mean square error is used as a loss function, and dynamic learning rate scheduling and an early stop mechanism are combined to improve convergence stability and model training efficiency.

The invention belongs to the technical field of communication, and particularly relates to a multipath channel DOA estimation method based on a heterogeneous attention double-branch neural network, and the method comprises the steps: constructing a training set, a verification set and a test set through stratified sampling; introducing a multi-head attention mechanism to construct a heterogeneous attention double-branch neural network for realizing intelligent DOA estimation; designing a frequency weighted loss function as a loss function of neural network model training; a dynamic attention mechanism and an early stop mechanism are designed in a model training process to prevent an overfitting phenomenon. According to the method, the problem of serious DOA estimation model overfitting in a small sample and data non-uniform real acquisition data scene is effectively relieved, and the DOA estimation accuracy is improved.

The invention provides a visual-text cross-modal panda behavior recognition method based on an attention mechanism, and relates to the technical field of attention mechanisms, and the method comprises the steps: inputting a multi-modal data set into an initial model, extracting panda behavior features, introducing a customized cross-modal attention mechanism to achieve feature alignment, and enhancing the visual and text feature interaction depth; the bottleneck of insufficient semantic fusion is broken through; constructing video frame sequence-text description paired samples based on features, carrying out bidirectional matching learning in a unified embedding space through a cross-modal representation network, optimizing parameters through symmetric cross entropy loss, curing a model in combination with a verification and early stop mechanism, capturing video time sequence information, and solving the problem of incomplete behavior dynamic representation; and finally, the visual part of the target panda behavior data is optimized, a preprocessing strategy is adjusted according to quality parameters, an optimization result assists an attention mechanism to focus on key features, the current situation that preprocessing does not have a unified standard and cannot be fed back is improved, and finally the target behavior category is accurately recognized.

The invention discloses a Text2SQL large model fine tuning method and device based on group relative strategy optimization, equipment and a medium, and relates to the technical field of interaction between natural languages and databases. The method comprises the following steps: S1, constructing a training data set, loading a pre-trained Text2SQL base model as a strategy network, initializing the strategy network, and obtaining an initial strategy; s2, performing grouping sampling on the training data set to obtain a candidate set; and S3, according to the candidate set, performing multi-dimensional comprehensive evaluation on the candidate SQL of the candidate set by adopting a weighted combination reward function to obtain relative advantages. And S4, loss is calculated according to the initial strategy and the relative advantage, and parameter updating is carried out. And S5, iterating through a performance monitoring system, an adaptive training strategy, a convergence judgment mechanism and an early stop and model selection strategy to obtain a trained model. According to the method, efficient learning and generation of the model on complex SQL query are realized, and the accuracy and the performability of a Text2SQL task are improved.

The invention discloses a work order semantics and load curve linked metering anomaly positioning method, which relates to the technical field of electric power metering, and is characterized in that an OBIS event is used as a strong anchor, work order time is used as a soft anchor, and acquisition time is used as a sampling anchor to construct a unified time axis, and an optimal window, an operation mask, and machine account confidence and time alignment mapping are output. Decomposing the load curve into trend, period and disturbance residual errors in the optimal window, and decoding OBIS events and work order elements in combination with the morphological dictionary and neighborhood consistency to form three-domain evidence vectors; and fusing the reconstructed residual error, the density rare degree and the single-class interval to generate a candidate list, and solidifying an evidence chain by mask shunting and early stop noise reduction. According to the method, events and semantics are taken as priori, forms are taken as likelihood, a virtual repair operator is introduced to perform anti-fact verification, non-technical loss can be diffused in a transformer area map to cut off an overhaul reconstruction edge, an evidence card write-back closed loop is generated, and recall, cause determination and positioning are improved.

The invention relates to the technical field of space structure design and parameter optimization, in particular to a neural network model-based reticulated shell structure node parameter automatic optimization method, which comprises the following steps of: acquiring an input-output data pair of a single-layer cylindrical reticulated shell constructed by aluminum alloy plate type nodes, the input parameters comprise design parameters and initial node parameters of the single-layer cylindrical reticulated shell. According to the method, node optimization data under different parameters are obtained through interaction of a genetic algorithm, finite element software and a programming tool, an example is supplemented to construct a training data set covering a common design parameter range of a project, and then a three-layer full-connection neural network model containing two hidden layers is trained; and the model precision is ensured by matching an early stop strategy and a learning rate attenuation strategy. During application, target single-layer cylindrical reticulated shell design parameters are input, the model can automatically output optimized node parameters, repeated iteration and manual intervention of a traditional method are not needed, optimization time consumption is greatly shortened, and the multi-working-condition batch optimization requirement is efficiently met.

The invention provides a communication radiation source individual identification method based on time domain and time sequence coding constellation diagram feature fusion. The method is used for realizing stable identification of different transmitters under non-ideal conditions of carrier frequency offset, noise interference and the like. The method comprises the following steps: firstly, carrying out normalization and fixed-length interception on a received I / Q signal, and calculating an oversampling multiple according to a symbol rate and a time resolution; then, a time domain and time sequence coding constellation diagram double-branch feature extraction network is constructed, a time domain structure feature is extracted from a time domain branch by adopting a one-dimensional modified ResNet18, and a constellation diagram with time information is generated from a time sequence coding constellation diagram branch through a time sequence coding constellation module and the feature that the phase changes along with time is extracted; and carrying out joint modeling on the two types of features through a trainable weighted fusion mechanism. In the training stage, the performance of the model is improved by adopting cross entropy loss, an Adam optimizer, learning rate scheduling and an early stop strategy. In the reasoning process, input signals are classified by using the model weight with the highest verification accuracy, and a communication radiation source individual identification result is output. Experimental results show that the method can effectively enhance the frequency offset robustness and generalization ability of the model, realizes high-precision identification of the individual communication radiation source, and is suitable for the fields of security access of the Internet of Things, spectrum monitoring, electronic countermeasures and the like.

The invention provides a ship ballast water system fault diagnosis model establishment method, belongs to the technical field of ship ballast water systems, and adopts a random noise data enhancement and maximum and minimum value standardization preprocessing method to improve data quality. Designing a CNN-I nformer hybrid architecture, fusing a probability sparse self-attention mechanism and distillation operation to extract time sequence features, and establishing a double-layer game optimization framework including an adaptive spiral flight sparrow search algorithm to perform joint optimization on model hyper-parameters; a softmax classifier is adopted to realize accurate identification of five states of sea chest filter filth blockage, ballast water pump bearing wear, valve blockage, pipeline leakage and system normal operation, and generalization ability and practicability of the model are ensured through an early stop strategy and multi-index performance evaluation. The technical problems that the ship ballast water system fault diagnosis precision is insufficient and the calculation complexity is too high are solved.

The invention discloses a hydropower station reservoir runoff intelligent forecasting method based on a large language model, and belongs to the technical field of hydrological forecasting, and the method comprises the following steps: S1, data collection and preprocessing; s2, constructing a BERT-Hydrology deep learning model for the runoff forecast of the reservoir of the hydropower station; s3, designing a weighted mixed loss function, and introducing an optimizer, learning rate scheduling and an early stop mechanism to improve the convergence speed and numerical stability of model training; and S4, establishing an evaluation system, performing rolling forecasting, and outputting an evaluation index and a significance test result. According to the intelligent hydropower station reservoir runoff forecasting method based on the large language model, the accuracy and the stability of runoff forecasting are remarkably improved by constructing the runoff forecasting model, excavating the hidden rule in the hydrological sequence and establishing the multi-step forecasting mapping, and the forecasting accuracy and the forecasting stability are improved. And technical support is provided for intelligent scheduling of a hydropower station reservoir and efficient management of water resources.

The invention belongs to the field of resource allocation, particularly relates to a multi-chain resource allocation technology based on a neural network, and provides the following scheme that the multi-chain resource allocation technology comprises a block chain multi-chain system and further comprises the following steps: S1, preparing training data by using a neural network training module: preparing the training data by using the neural network training module; s2, training a node optimization distribution module: dynamically adjusting network parameters through an adaptive optimization algorithm based on the training data set processed by the neural network training module, and controlling an iteration process by adopting an early stop mechanism and a learning rate attenuation strategy until a loss function is converged to a preset threshold value, according to the method, node attributes and global load information are dynamically fused through the graph neural network, intelligent optimal distribution of nodes is achieved, meanwhile, the over-smoothing problem caused by multi-layer stacking can be avoided while light weight of the model is guaranteed, and through intelligent decision making and dynamic adaptation capacity, the optimal distribution of the nodes is achieved. And the problems of resource mismatching, system bottleneck and expansibility are solved.

The invention relates to a multi-channel multi-mode space-time 3D convolution and Transform fused satellite-borne GNSS-R wave height inversion model, and aims to solve the problems that a traditional empirical model is poor in adaptability to complex oceans and existing deep learning cannot fully excavate space-time correlation. According to the method, three modules are innovatively fused for cooperative processing: firstly, a 3D CNN-ConvLSTM is utilized to extract local spatial-temporal characteristics of multichannel GNSS-R data and capture a dynamic time sequence; secondly, performing deep coding on the sea surface environment parameters through a multi-head attention mechanism of Transform, and establishing global dependence between features; and finally, realizing cross-modal fusion by adopting a weighted summation and feature splicing strategy. And through training of an Adam optimizer and control of an early stop strategy, optimization is carried out by taking a mean square error as a loss function. Experiments show that compared with a traditional model and a machine learning model, the method has the advantages that the significant wave height estimation error is reduced by 40-53%, the correlation coefficient reaches 0.84-0.91, the precision and generalization ability under the complex sea condition are remarkably improved, and reliable support is provided for ocean remote sensing monitoring and disaster early warning.

This invention discloses a deep learning-based method for predicting the phenological synchronization of specialized butterflies and their host plants, belonging to the interdisciplinary field of biodiversity conservation, ecological prediction, and deep learning technology. The method includes: collecting environmental climate variables, butterfly occurrence data, and phenological time-series data of their host plants; preprocessing the data to create training and validation sets; constructing a dual-label system with the day of year (DOY) of butterfly occurrence peaks (e.g., emergence peak, egg-laying peak) as the primary label and the normalized value of synchronization deviation as the secondary label; building a deep learning model with ecological weights, including an embedding layer, an LSTM layer, and a fusion layer; inputting a vector concatenated from daily or weekly climate time-series features and ecological dynamic features; and outputting the primary and secondary labels through a dual-output layer; designing a hybrid loss function that integrates the main regression loss, ecological synchronization constraint loss, and small-sample regularization loss; training the model using the Adam optimizer and implementing an early stopping mechanism; and validating and evaluating the model using regression accuracy and ecological synchronization accuracy indicators. The method includes modules for data acquisition, preprocessing, label construction, model training, prediction evaluation, and storage. This invention improves the accuracy of phenological synchronous prediction by embedding ecological dynamic feature weights and dual-label constraints, adapts to small sample scenarios, and provides technical support for the monitoring and protection management of rare butterfly species (such as the Golden Birdwing and the Chinese Tiger Swallowtail).

The invention discloses a dynamic arrangement and iterative optimization method for a skill map in the water conservancy field, and relates to the technical field of artificial intelligence and informatization. Comprising the steps of 1, performing memory-driven industry thematic problem rewriting, 2, performing hybrid intention recognition and tool calling routing, 3, performing dynamic skill graph arrangement based on graph calculation: constructing a water conservancy field skill graph by utilizing a graph calculation technology, representing skill nodes and subtasks as a directed acyclic graph (DAG), and performing dynamic skill graph arrangement based on graph calculation. The method comprises the following steps of: 1, selecting an intention recognition result, matching skill nodes and subtasks according to task requirements in combination with the routing strategy and the intention recognition result to dynamically generate an execution plan, 4, carrying out parameter self-diagnosis and multi-round clarification, and 5, carrying out ReAct closed-loop feedback and field iteration: adopting a ReAct normal form to combine model reasoning and actions to form multi-round closed-loop interaction, and carrying out field iteration on the multi-round closed-loop interaction. And adjusting a reasoning path according to feedback in each round of iteration, introducing an early stop mechanism, and terminating iteration in advance when a model execution plan generation result accords with expectation.

The invention relates to the field of electric power and data mining, in particular to a standard necessary patent prediction model construction method, prediction method and system, and the method comprises the steps: obtaining a training data set, the data in the training data set being a plurality of pieces of patent data with known judgment results; inputting the preprocessed data of each patent into a preset neural network model to respectively obtain a probability prediction value that the data of each patent is a standard necessary patent; optimizing parameters in a preset neural network model on the basis of the difference between the probability prediction value that each patent data is the standard necessary patent and the corresponding known judgment result until a preset training stopping condition is met, and obtaining a standard necessary patent prediction model; wherein the preset training stopping condition comprises the steps that the number of training rounds of the preset neural network model is controlled through an early stopping method, and when the loss of a certain round of training of the preset neural network model is larger than the loss of a corresponding last round of training, training of the preset neural network model is stopped.

The invention provides a multi-agent combined model automatic training method and a related product. According to one specific mode of the method, a decision-making agent corresponding to a target business scene is utilized to determine a lower-round hyper-parameter optimization execution agent for controlling a lower-round hyper-parameter optimization scheme of a target machine learning model; and based on the generated parameter adjustment record, a next-round hyper-parameter optimization execution agent is utilized to execute corresponding control operation so as to stop training of the target machine learning model, or replace a hyper-parameter search space of the target machine learning model, or determine next-round hyper-parameter configuration of the target machine learning model. And training the target machine learning model according to the next-round hyper-parameter configuration. Therefore, a traditional Bayesian hyper-parameter optimization method is intelligently transformed; the model training frequency is reduced; the model construction efficiency is improved; the model performance is improved; and self-adaptive model training is realized by automatically judging early stopping and continuing training.

The invention relates to the field of compressible turbulence, in particular to a method for predicting a beam expansion curve in compressible turbulence based on an LSTM (Long Short Term Memory) model, which comprises the following steps of: extracting multi-dimensional features of an input beam expansion front section curve, and converting the multi-dimensional features into an LSTM model adaptive format; a depth LSTM model is constructed; customizing a callback function based on a cosine annealing principle, and periodically adjusting a learning rate according to the callback function; integrating early stop, optimal model storage and learning rate scheduling callback strategies, dynamically adjusting the learning rate of the LSTM model, and optimizing the LSTM model; and the LSTM model sequentially executes a main training whole process, restores a real scale through reverse normalization, and outputs a prediction result of the beam expansion curve in the compressible turbulence. According to the method, the LSTM neural network model can be stably established, the model stability is good, the prediction result of the beam expansion curve in compressible turbulence is quantified, and the prediction result is accurate.

The invention discloses a data processing method and device, electronic equipment and a storage medium, and relates to the technical field of computers. A specific embodiment of the method comprises the following steps: receiving an experiment trigger instruction, and obtaining corresponding historical sample data to determine experiment sample data and contrast sample data; querying an early stop index type, determining an early stop index based on the early stop index type, and calculating a corresponding early stop threshold value; executing an experimental operation program on the experimental sample data to obtain an experimental result, and calculating a first conversion quantity and a second conversion quantity of the early stop index in the experimental result and the contrast sample data according to the early stop threshold value to determine a change parameter of the early stop index; and in response to the change parameter meeting the early stop condition, stopping the experimental operation program. According to the implementation mode, the problem that time and resource waste is easily caused by a row mode that the experiment result is analyzed only after the execution of the preset sample size is finished or the experiment is executed for the preset time period in the experiment can be solved.

The invention discloses a synonymous mutation harmfulness prediction method based on a gene language model, and the method comprises the steps: (1) data obtaining and screening: employing a data set existing in a laboratory, and guaranteeing the data reliability and task pertinence; (2) carrying out feature annotation: respectively carrying out embedded feature extraction on the reference sequence and the mutation sequence by utilizing four representative gene language models GPN-MSA, Hyena DNA, SpliceBERT and Caduceus; (3) feature processing and fusion: performing further extraction on GPN-MSA and SpliceBERT embedding by adopting dynamic convolution and bidirectional LSTM (Long Short Term Memory); extracting the embedding of the Hyena DNA by adopting TextCNN (Convolutional Neural Network); linear dimensionality reduction is carried out on Caduceus embedding, and then splicing and dimensionality reduction are carried out on the four features; (4) model training: training in a five-fold cross validation mode, and preventing over-fitting through an early stop mechanism to obtain an optimal weight; and (5) model prediction: performing harmfulness prediction by taking 0.5 as a threshold value. According to the method, the complementary information of the multi-gene language model is fused, the prediction precision and robustness of the synonymous mutation harmfulness are remarkably improved, and a new technical support is provided for research on the synonymous mutation harmfulness.

The application discloses a kind of fusion attention mechanism and the industrial bearing vibration time series signal fault prediction method and system of LSTM.This method includes: collecting bearing vibration signal and carrying out filtering, noise reduction and normalization preprocessing;BiLSTM and coordinate attention mechanism combined deep learning model is constructed to extract bidirectional time series features and enhance key fault features;Multi-objective composite loss function and Adam optimizer are used for model training, and early stopping mechanism is introduced to prevent overfitting;Real-time vibration signal is used to identify fault type and degree evaluation using the trained model, and multi-scale spectral kurtosis features and nonlinear dynamics parameters are used for quantitative analysis;Finally, output fault diagnosis report, and trigger multi-level early warning based on adaptive threshold value.The application can realize high-precision, high-reliability bearing fault prediction and health state evaluation, and is suitable for industrial equipment intelligent operation and maintenance.

The invention belongs to the field of machine learning, and discloses a reasoning acceleration method and system for a diffusion large language model, which are used for distinguishing a calculation activity area and a stability area of the diffusion large language model; in the subsequent diffusion step, only the increment output by the active area is calculated, and the increment is fused with the calculation result of the stable area to obtain an intermediate result; calculating a token-level generation entropy based on the intermediate calculation result, thereby identifying a decoded token set to achieve an early stop; and finally, dynamically adjusting hardware calculation scheduling. According to the method, the temporal locality of a calculation unit between continuous diffusion steps is utilized, repeated calculation of complete forward transmission is avoided through a cache stable region, and redundant calculation and memory bandwidth waste are reduced; the token level generation entropy value dynamically monitors the convergence state of the token, the reasoning process is submitted to be terminated, and excessive denoising and invalid calculation caused by fixed-step reasoning are avoided; dynamic scheduling adapts to a sparse calculation mode and irregular memory access required by the model, and the problem of mismatching of hardware and software is relieved.

The invention discloses a cross-regional edge node congestion collaborative early warning method based on federated learning. The method comprises the steps of data set construction, federated learning framework initialization, local model training, global model aggregation, congestion detection and congestion early warning. The invention relates to the technical field of cross-regional edge network intelligent monitoring, in particular to a cross-regional edge node congestion collaborative early warning method based on federated learning, which comprises the following steps of: initializing a sequence model at an edge node, constructing local loss combined with binary cross entropy and time sequence smooth punishment, and stably training by Adam and early stop; through weighted average aggregation disturbance parameters based on data volume and loss, performance degradation after fusion is prevented by adopting momentum fusion and a rollback threshold value; and meanwhile, an attention mechanism is embedded to realize joint prediction of probability, severity and duration, cross-regional propagation parameters of the graph neural network are utilized, loss is verified, and invalid sharing is rejected, so that prediction accuracy and timeliness are improved, and node congestion is effectively relieved.

The invention belongs to the field of knowledge cognition level mining and adaptive teaching systems, and provides a personalized cognition level mining method based on subjective and objective test question collaborative modeling. Obtaining learner-test question interaction data, constructing a knowledge point vector according to the Q matrix, and generating an initial mastering vector based on the student ID; three types of personalized parameters including slip, guess and difficulty are introduced, and the difficulty and the mastery degree are modulated and then spliced with the Q matrix to form a feature sequence; a bidirectional LSTM is adopted to extract time sequence dependence, a knowledge state input update network is combined, and state estimation is dynamically generated; and splicing with a test question Q vector, inputting into a score prediction network, and fusing slip and guess to construct an improved IRT function. Training is supervised by using historical answer records, prediction errors are minimized, an early stop mechanism is introduced, objective question accuracy and subjective question mean square errors are output, multi-dimensional evaluation is realized, and prediction performance and universality are improved.

The application relates to a SO4 based on deep cross network (DCN V2) feature enhancement and integrated learning 2‑ The concentration driving factor analysis method comprises the following steps: collecting time series data of atmospheric pollutant concentration and meteorological parameters in a target area, and performing pretreatment and quality control; a DCN V2 model integrating a linear cross network, an exponential cross network and a deep neural network is constructed to realize nonlinear high-order cross feature enhancement of original features; a two-layer Stacking integrated learning model taking XGBoost, LightGBM and CatBoost as heterogeneous base models and linear regression as a meta model is constructed, and global optimization is performed through Bayesian optimization; and the SHAP method is applied to quantify the marginal contribution and direction of each feature, and identify the key driving factors of SO4 2‑ concentration. The application integrates optimization strategies such as GPU acceleration, early stopping mechanism, automatic task detection, parallel training and backup training, realizes the automation, high efficiency and reproducibility of the analysis process, and guarantees the physical interpretability of the driving factor analysis, thereby providing scientific and technological support for atmospheric pollution prevention and decision-making.