Small Sample Data Research Articles

Research on Node Fault Diagnosis in Wireless Sensor Networks Based on a Constrained Data Maximum Entropy BN Parameter Learning Algorithm

Currently, power communication wireless sensor networks (WSNs) exhibit the characteristics of uncertainty and complexity. Furthermore, the application environment is more complex, resulting in nodes damaging easily, power communication breakdown, and economic losses. Dynamic monitoring of power communication WSN nodes is necessary. To this end, a constraint data maximum entropy Bayesian network (BN) parameter learning algorithm is used to solve the problem of small data sample size in monitoring and improve the quality of power communication WSN node fault diagnosis. After the latest verification, it is found that the correct rate of WSN node fault diagnosis under small data set is as follows: in normal conditions, the fault rate is 96%, the sensor fault is 94%, the power supply fault is 100%, the wireless communication fault is 90%, and the processor fault is 91%. It can be seen that even under small sample data sets, better diagnostic accuracy can be brought.

Research on the Prediction Model of Coal Spontaneous Combustion Hazard Level Based on IGWO-GRNN

ABSTRACT To accurately predict the risk level of spontaneous coal combustion, we introduce an improved grey wolf optimization strategy (IGWO) to optimize the generalized regression neural network (GRNN), thereby establishing the IGWO-GRNN coal spontaneous combustion risk level prediction model. The superiority of the IGWO algorithm over the grey wolf optimization algorithm (GWO), whale optimization algorithm (WOA) and particle swarm optimization algorithm (PSO) was verified. Fifty sets of sample data of spontaneous coal combustion in the mining area were selected as the research object, and the data were randomly divided into the training set and the test set according to the ratio of 8:2. The prediction results of the IGWO-GRNN prediction model were obtained and compared with those of the GWO-GRNN, GRNN, BP (Backpropagation), and GRU (Gated Recurrent Unit) models, and the absolute value of the relative error was introduced as an evaluation index of the model prediction accuracy. The results demonstrate that IGWO outperforms GWO, WOA, and PSO in terms of optimization efficiency, with the IGWO-GRNN prediction model yielding the most accurate results. Specifically, the maximum absolute value of the relative error does not exceed 4.27%, and the average relative error is only 3.29%, indicating a substantial improvement over the other models. Therefore, this study achieves high-precision prediction under small-sample data, which provides an important reference value for the small-sample prediction problem in the mining field.

Phenomenological Modeling of Antibody Response from Vaccine Strain Composition

The elicitation of broadly neutralizing antibodies (bnAbs) is a major goal of vaccine design for highly mutable pathogens, such as influenza, HIV, and coronavirus. Although many rational vaccine design strategies for eliciting bnAbs have been devised, their efficacies need to be evaluated in preclinical animal models and in clinical trials. To improve outcomes for such vaccines, it would be useful to develop methods that can predict vaccine efficacies against arbitrary pathogen variants. As a step in this direction, here, we describe a simple biologically motivated model of antibody reactivity elicited by nanoparticle-based vaccines using only antigen amino acid sequences, parametrized with a small sample of experimental antibody binding data from influenza or SARS-CoV-2 nanoparticle vaccinations. Results: The model is able to recapitulate the experimental data to within experimental uncertainty, is relatively insensitive to the choice of the parametrization/training set, and provides qualitative predictions about the antigenic epitopes exploited by the vaccine, which are testable by experiment. For the mosaic nanoparticle vaccines considered here, model results suggest indirectly that the sera obtained from vaccinated mice contain bnAbs, rather than simply different strain-specific Abs. Although the present model was motivated by nanoparticle vaccines, we also apply it to a mutlivalent mRNA flu vaccination study, and demonstrate good recapitulation of experimental results. This suggests that the model formalism is, in principle, sufficiently flexible to accommodate different vaccination strategies. Finally, we show how the model could be used to rank the efficacies of vaccines with different antigen compositions. Conclusion: Overall, this study suggests that simple models of vaccine efficacy parametrized with modest amounts of experimental data could be used to compare the effectiveness of designed vaccines.

Study on corrosion monitoring and assessment method of reinforced concrete based on multi-sensor fusion

Purpose The purpose of this paper is to investigate the accurate monitoring and assessment of steel bar corrosion in concrete based on deep learning multi-sensor information fusion method. The paper addresses the issue of traditional corrosion assessment models relying on sufficient data volume and low evaluation accuracy under small sample conditions. Design/methodology/approach A multi-sensor integrated corrosion monitoring equipment for reinforced concrete is designed to detect corrosion parameters such as corrosion potential, current, impedance, electromagnetic signal and steel bar stress, as well as environmental parameters such as internal temperature, humidity and chloride ion concentration of concrete. To overcome the small amount of monitoring data and improve the accuracy of evaluation, an improved Siamese neural network based on the attention mechanism and multi-loss fusion function is proposed to establish a corrosion evaluation model suitable for small sample data. Findings The corrosion assessment model has an accuracy of 98.41%, which is 20% more accurate than traditional models. Practical implications Timely maintenance of buildings according to corrosion evaluation results can improve maintenance efficiency and reduce maintenance costs, which is of great significance to ensure structural safety. Originality/value The corrosion monitoring equipment for reinforced concrete designed in this paper can realize the whole process of monitoring inside the concrete. The proposed corrosion evaluation model for reinforced concrete based on Siamese neural network has high accuracy and can provide a more accurate assessment model for structural health testing.

Toward secure industrial internet of behaviours: a federated learning-based lightweight human behaviour recognition method with selective state space models

Human behaviour recognition is one of the most fundamental tasks in Industrial Internet of Behaviour (IIoB) and is crucial for the safe and reliable IIoB. Existing methods lacks adaptability and transferability. In addition, there is a data isolation problem among different users. Therefore, there is an urgent requirement to construct a secure and adaptive human behaviour recognition model in IIoB without violating the privacy of users. Mamba, a structured state space model that integrates a selection mechanism and a scan module, is used for time series modelling tasks. To tackle the aforementioned problems, an Federated Learning-based lightweight human behaviour recognition model with selective state space models is proposed. First, we design a human behaviour recognition model integrating Mamba and residual structure to achieve lightweight and secure human behaviour modelling. In addition, considering data privacy and training efficiency, a decentralised dynamic FL framework is designed to achieve lightweight and secure model collaborative training, including: initial selection of source users, model aggregation strategy based on dynamic weighting, and fine-tuning module based on small-sample data, to improve the training efficiency of the model and the accuracy of human behaviour recognition. Extensive experiments are conducted to prove the superior performance of the proposed method.

A droplet state prediction method for inkjet printing under small sample conditions based on the two-stage TrAdaBoost.R2 algorithm

Inkjet printing is regarded as a new generation of green intelligent manufacturing technology. However, precise control of the state of droplets in inkjet printing is critical and costly. In this study, a voltage-driven signal-based ink droplet state prediction model suitable for a small sample data environment is proposed. Seven public ink datasets and self-collected experimental data are grouped together, and the decision tree, AdaBoost.R2, and two-stage TrAdaBoost.R2 algorithms are used, respectively. The two-stage TrAdaBoost.R2 model shows excellent prediction performance in the case of insufficient data by transferring knowledge of different inks. In particular, it achieves mean square errors of 0.00219 for droplet volume predictions and 0.0645 for velocity predictions. Similarly, the mean absolute percentage errors are 0.91% for droplet volume and 1.37% for velocity. Furthermore, the two-stage TrAdaBoost.R2 model demonstrates strong generalization ability and adaptability to maintain high prediction accuracy under different conditions. The results indicate that the two-stage TrAdaBoost.R2 effectively mitigates the prediction performance issues caused by data scarcity, paving the way for technological advancements in the field of inkjet printing.

Complex product network change prediction method based on GANs with small sample data

Complex product network change prediction method based on GANs with small sample data

О ВОЗМОЖНОСТИ ПОВЫШЕНИЯ СТАТИСТИЧЕСКОЙ ЗНАЧИМОСТИ РЕГРЕССИОННЫХ ЗАВИСИМОСТЕЙ В ЗАДАЧАХ ОБРАБОТКИ ЭКСПЕРИМЕНТАЛЬНЫХ ВЫБОРОК МАЛОГО ОБЪЕМА

The article considers the problem of increasing the reliability and statistical significance of regression models of research objects built on small experimental data samples. Insufficient amount of experimental data forces the researcher to use linear models with a minimum number of variable factors, however, even with such a choice of the type of model, insufficient statistical significance of parameter estimates excludes the possibility of using it for reliable forecasting of changes in the explained variables. In order to expand the possibility of choosing the type of model at the specification stage and to increase the statistical significance of its parameter estimates, it is proposed to expand the volume of experimental data using a statistical model of the object of study, built on the basis of a generative adversarial neural network. When training on a small sample obtained during an experimental study of the object, the generator of a conditional generative adversarial network generates data clusters with centroids corresponding to the points of the training (experimental) sample. The results of the analysis of the data of a physical experiment are presented, confirming its main provisions.

Research on Small-Sample Credit Card Fraud Identification Based on Temporal Attention-Boundary-Enhanced Prototype Network

The Nielsen Report points out that credit card fraud caused business losses of USD 28.65 billion globally in 2019, with the US accounting for more than one-third of the high share, and that insufficient identification of credit card fraud has brought about a serious loss of financial institutions’ ability to do business. In small sample data environments, traditional fraud detection methods based on prototype network models struggle with the loss of time-series features and the challenge of identifying the uncorrected sample distribution in the metric space. In this paper, we propose a credit card fraud detection method called the Time-Series Attention-Boundary-Enhanced Prototype Network (TABEP), which strengthens the temporal feature dependency between channels by incorporating a time-series attention module to achieve channel temporal fusion feature acquisition. Additionally, nearest-neighbor boundary loss is introduced after the computation of the prototype-like network model to adjust the overall distribution of features in the metric space and to clarify the representation boundaries of the prototype-like model. Experimental results show that the TABEP model achieves higher accuracy in credit card fraud detection compared to five existing baseline prototype network methods, better fits the overall data distribution, and significantly improves fraud detection performance. This study highlights the effectiveness of open innovation methods in addressing complex financial security problems, which is of great significance for promoting technological advancement in the field of credit card security.

Adaptive Multicore Dual-Path Fusion Multimodel Extraction of Heterogeneous Features for FAIMS Spectral Analysis.

With the increasing application scenarios and detection needs of high-field asymmetric waveform ion mobility spectrometry (FAIMS) analysis, deep learning-assisted spectral analysis has become an important method to improve the analytical effect and work efficiency. However, a single model has limitations in generalizing to different types of tasks, and a model trained from one batch of spectral data is difficult to achieve good results on another task with large differences. To address this problem, this study proposes an adaptive multicore dual-path fusion multimodel extraction of heterogeneous features for FAIMS spectral analysis model in conjunction with FAIMS small-sample data analysis scenarios. Multinetwork complementarity is achieved through multimodel feature extraction, adaptive feature fusion module adjusts feature size and dimension fusion to heterogeneous features, and multicore dual-path fusion can capture and integrate information at all scales and levels. The model's performance improves dramatically when performing complex mixture multiclassification tasks: accuracy, precision, recall, f1-score, and micro-AUC reach 98.11%, 98.66%, 98.33%, 98.30%, and 98.98%. The metrics for the generalization test using the untrained xylene isomer data were 96.42%, 96.66%, 96.96%, 96.65%, and 97.60%. The model not only exhibits excellent analytical results on preexisting data but also demonstrates good generalization ability on untrained data.

Surface inspection of annular ceramic metal coating based on small sample data augmentation and deep learning

Surface inspection of annular ceramic metal coating based on small sample data augmentation and deep learning

The fusion method based on small-sample aerodynamic thermal and force data

At present, high-fidelity data are expensive to acquire. When fusing limited high-fidelity data, the small-sample size introduces problems such as missing information and sample bias, which leads to overfitting of the results and accuracy degradation. In this paper, we propose a small-sample aerodynamic data fusion method based on deep neural networks. The method applies semi-supervised learning for model construction using multi-fidelity aerodynamic thermal and force data. The initial model is trained with both labeled and unlabeled data by an improved flexible loss function. Using unlabeled data as a soft constraint combined with semi-supervised learning enables the model to perform better with small-sample data. This article investigates the ONERA (National Office for Aerospace Studies and Research) M6 wing surface pressure distributions at different airfoil spread coordinates and verifies the applicability of the proposed method by reducing the proportion of high-fidelity data in the training and test datasets. The proposed method is then applied to the prediction of aerothermal data on the surface of a blunt bicone. The results show that, using a small-sample high-fidelity dataset, the proposed method can predict the surface pressure distribution and surface aerodynamic heat distribution of the aircraft relatively well. As the volume of high-fidelity data decreases, the proposed method outperforms other methods.

Improving the Reliability of Biometric Authentication Processes Using a Model for Reducing Data Drift

Modern complexes providing biometric identification face several problems, such as information drift caused by the variability of facial patterns, voice timbres, and current states. Information drift can characteristically exhibit short-term (subjects' states have changed) or long-term changes. Simultaneously, the developed trusted systems should not have the properties of explainable AI to prevent the possibility of intruders, based on understanding the system behavior to perform actions to hack the system. This paper's objective is to improve the reliability of biometric authentication by increasing the informativity of the classified images by transforming the correlations between the information features using the Bayes-Minkowski measure. The paper puts forth the proposition of employing neuroimmune models that are founded upon the principles of both acquired and innate immunity, with an analogy to the natural immune system. In addition, the authors propose to analyze correlations between information features instead of the features themselves. To reduce the influence of data drift, the authors suggest using adaptive learning with a teacher and reinforcement, which helps to work even with small and unrepresentative data samples. The proposed algorithm demonstrates a high degree of accuracy, as evidenced by its equal error rate (EER), and is particularly well-suited to feature recognition tasks due to its adaptive model. The test results have shown that the proposed solutions increase the level of security of personal data and improve the reliability of biometric authentication against fraudulent actions of intruders, including approaches based on adversarial algorithms. The integration of the immune structure into the authentication system enables the algorithm to remain stable even when presented with a limited number of samples. The proposed algorithm mitigates the impact of data drift on the authentication outcome. Doi: 10.28991/ESJ-2024-08-06-018 Full Text: PDF

Fault diagnosis of HVCB via the subtraction average based optimizer algorithm optimized multi channel CNN-SABO-SVM network

The mechanical fault diagnosis of HVCB is important to ensure the stability of electric power systems. Aiming at the problem of poor diagnostic performance of deep learning methods under limited samples, this paper proposes an HVCB operating mechanism fault diagnosis model (multi-channel CNN-SABO-SVM, MCCSS) based on multimodal data fusion features and Subtraction-Average-Based Optimizer (SABO). This model extracts and fuses features from the input two-dimensional data using a multi-channel CNN network and then uses the multimodal data fusion features to diagnose HVCB faults. Additionally, the SVM is used instead of the Softmax classifier to classify the fused features of vibration and sound, compensating for the poor diagnostic performance and generalization ability of the CNN network in small sample data scenarios. To further enhance the fault diagnosis performance of the SVM, the SABO is introduced for hyperparameter optimization of the SVM classifier. An HVCB fault test platform was established to train and test the model with limited data. The experimental results show that, compared with the multi-channel CNN-SVM and the CNN model based on unimodal signals, the proposed multi-channel CNN-SABO-SVM model improves the accuracy by 2.66% and 10.66%, respectively, and effectively addresses the challenge of circuit breaker fault diagnosis with limited samples.

Cross-operating-condition fault diagnosis of a small module reactor based on CNN-LSTM transfer learning with limited data

Fault samples of Small Modular Reactor (SMR) are scarce and the data distribution cross operating conditions are different, resulting in poor generalization performance of fault diagnosis model. In this paper, Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) are integrated and a novel CNN-LSTM transfer learning method is proposed for fault diagnosis of SMR at different power levels, which is used to improve the deep extraction capability of spatial and temporal features of fault data. Based on small sample data and feature visualization technology, five different feature transfer strategies are compared to obtain the best generalization performance of the cross-operating condition fault diagnosis model. In addition, the effects of different network structures and hyperparameter combination settings of the CNN-LSTM model on the convergence speed and accuracy of transfer learning are analyzed to optimize the design of SMR fault diagnosis model. The results show that the CNN-LSTM method has stronger fault feature extraction capability than other deep learning methods. The simulation results also show that the proposed transfer learning strategy and the optimized network structure and hyperparameters could significantly enhance the generalization performance and accuracy of SMR fault diagnosis under different operating conditions with few labeled samples.

Pipeline leak detection based on generative adversarial networks under small samples

During the actual industrial process, it is challenging to obtain samples of oil and gas pipeline leaks resulting in the scarcity of training data samples suitable for fault diagnosis. In order to tackle this issue, this paper suggests a method a Recursive Generalization self-attention generative adversarial network (RAGAN) aided by Wasserstein distance with gradient penalty. The initial step involves applying the short time Fourier transform to the acoustic signal of oil and gas pipeline leakage, treating it as real data. Subsequently, both the real data and random noise following a Gaussian distribution are fed into the generator. The output is utilised as a pseudo sample. The Wasserstein distance of the distribution of real data and fake samples is introduced as a loss term in the discriminator, and a gradient penalty is added. Finally, the network optimizes the parameters through back propagation until Nash equilibrium. PSNR and SSIM are used as sample reliability evaluation. The results show that the fake samples have high similarity with the real samples, which can be used to expand small sample data. Moreover, extending pseudo samples to small sample data sets can effectively improve the performance of fault diagnosis.

Reliability analysis of DDV and SPM in BOP-based Weibull model and expand fault dataset using deep learning

The hydraulic systems of deepwater Blowout Preventers (BOPs) are crucial for ensuring the reliability and safety of offshore oil and gas extraction operations. A functional failure can lead to uncontrolled blowouts, resulting in casualties and significant economic losses on the rig. The Direct Drive Valve (DDV) and the Subsea Plated Mounted (SPM) valve are key components that help maintain the proper functioning of the hydraulic system in deepwater BOPs. This study begins by utilizing the Weibull analysis method to assess the reliability of the DDV and SPM valves using limited fault data samples. To enhance the accuracy of predictions, Weibull parameters are estimated through various methods, including Maximum Likelihood Estimation (MLE), Least Squares Estimation (LSE), and a combination of Correlation Coefficient Optimization with Support Vector Regression (CCO + SVR).Given the challenges in gathering extensive fault data for DDV and SPM valves—due to complex subsea environments, cost constraints, time limitations, and other factors—this study proposes a method employing a Back Propagation Neural Network (BPNN) model to augment the limited fault data samples. To ensure the reliable operation of the DDV and SPM valves, preventive maintenance cycles are established at 2840 and 7550 operations, respectively. At the same reliability level, as the number of operational cycles increases, the remaining service life of the valves gradually decreases, leading to a higher probability of failure over a shorter timeframe. The Mean Remaining Life (MRL) of the DDV and SPM valves, corresponding to different operational times, is analyzed, providing essential reference points for their usage and maintenance. When the extended data sample is utilized for reliability evaluation, the reliability characteristics of the small fault data samples are effectively reflected, and the parameter prediction error remains low. This indicates that the extended data sample is more suitable for reliability evaluation.

Model of Dynamic Mechanical Parameters and Vibration Analysis in Hot Rolling

As the core equipment of metal shaping processing, the vibration problem of rolling mill seriously affects the product quality and production efficiency. To address the problem of the limited prediction of mill vibration due to small sample data in non-steady states, we propose to predict dynamic mechanical parameters (rolling force and rolling torque) based on the TCN-LSTM step strategy transfer model. The prediction accuracies of the TCN-LSTM step strategy transfer model under 1000 sets of training data reach 95.8% and 93.2%, respectively, and at the same time, it saves the time of training and regulation of the deep learning model and can better meet the needs of online prediction. The horizontal-torsion hot rolling vibration model is further established to quantitatively analyze the relationship between process parameters and mill vibration. The experimental results show that with the vibration suppression measures of 10% reduction of the rolling speed and 10% reduction of the entrance thickness, the horizontal vibration displacement amplitude is reduced from 0.687 × 10−4 m to 0.229 × 10−4 m, and the rotational displacement amplitude is reduced from 0.0273 m to 0.0117 m, which effectively suppresses the vibration of the mill and improves the stability of the rolling process.

Dealing with small samples in disability research: Do not fret, Bayesian analysis is here.

Small sample sizes are a common problem in disability research. Here, we show how Bayesian methods can be applied in small sample settings and the advantages that they provide. To illustrate, we provide a Bayesian analysis of employment status (employed vs. unemployed) for those with disability. Specifically, we apply empirically informed priors, based on large-sample (N = 95,593) July 2019 Current Population Survey (CPS) microdata to small subsamples (average n = 26) from July 2021 CPS microdata, defined by six specific difficulties (i.e., hearing, vision, cognitive, ambulatory, independent living, and self-care). We also conduct a sensitivity analysis, to illustrate how various priors (i.e., theory-driven, neutral, noninformative, and skeptical) impact Bayesian results (posterior distributions). Bayesian findings indicate that people with at least one difficulty (especially ambulatory, independent living, and cognitive difficulties) are less likely to be employed than people with no difficulties. Overall, results suggest that Bayesian analyses allow us to incorporate known information (e.g., previous research and theory) as priors, allowing researchers to learn more from small sample data than when conducting a traditional frequentist analysis. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Privacy-Preserving Incipient Fault Identification in Distribution Networks Under Small Sample and Imbalanced Data Distribution Conditions

Privacy-Preserving Incipient Fault Identification in Distribution Networks Under Small Sample and Imbalanced Data Distribution Conditions