Multi-domain Features Research Articles

An ensemble deep learning model for classification of students as weak and strong learners via multiparametric analysis

Academic data predictions are significantly important for improving the overall education system's effectiveness by providing early identification of weak students and personalized learning strategies. This paper proposes a deep learning model for the identification of weak and strong students using ensemble learning and multiparametric analysis. It combines several machine learning algorithms, including Naive Bayes, Support Vector Machines, Multi-Layer Perceptron, and Logistic Regression using an ensemble learning approach to enhance the model’s performance. Additionally, a custom 1D Convolutional Neural Network (CNN) is designed for classification. It utilizes multiparametric analysis to identify weak and strong students considering various parameters such as age, academic performance, location, and online learning behavior. The evaluation results indicate the performance of the proposed model has been improved in comparison to MLA FIS, SHMM, and DRL by 16.5%, 5.5%, and 2.4%, in terms of precision, 16.4%, 6.5%, and 3.5 % in terms of accuracy and 10.4%, 2.5% and 6.5% in terms of recall. These improvisations described that the model is efficient for multidomain feature extraction, ensemble classification, and high-variance feature selection, which result in a deeper understanding of student performance.

Single-Channel Sleep EEG Classification Method Based on LSTM and Hidden Markov Model

Background: The single-channel sleep EEG has the advantages of convenient collection, high-cost performance, and easy daily use, and it has been widely used in the classification of sleep stages. Methods: This paper proposes a single-channel sleep EEG classification method based on long short-term memory and a hidden Markov model (LSTM-HMM). First, the single-channel EEG is decomposed using wavelet transform (WT), and multi-domain features are extracted from the component signals to characterize the EEG characteristics fully. Considering the temporal nature of sleep stage changes, this paper uses a multi-step time series as the input for the model. After that, the multi-step time series features are input into the LSTM. Finally, the HMM improves the classification results, and the final prediction results are obtained. Results: A complete experiment was conducted on the Sleep-EDFx dataset. The results show that the proposed method can extract deep information from EEG and make full use of the sleep stage transition rule. The proposed method shows the best performance in single-channel sleep EEG classification; the accuracy, macro average F1 score, and kappa are 82.71%, 0.75, and 0.76, respectively. Conclusions: The proposed method can realize single-channel sleep EEG classification and provide a reference for other EEG classifications.

Small sample composite fault diagnosis of hydraulic bearings based on improved VME algorithm and mRVM

The working environment of rolling bearings is complex. Once a fault occurs, various parts will affect each other and produce a compound fault. Traditional methods often use signal separation algorithms to separate different types of signals for fault diagnosis, but it is difficult to analyze specific faults efficiently and accurately. To solve this problem, this paper combines variational mode decomposition (VMD), Laplace energy index (LE) and variational mode extraction (VME) for signal extraction. Multi-class relevance vector machine (mRVM) and DS evidence theory are used for intelligent fault diagnosis, focusing on the context of small sample data. First, the VMD-LE-VME method is used to extract effective fault information from the fault signal and obtain multi-domain features. Then, the multi-domain features are input into mRVM for fault identification. Finally, the classification results are fused through DS evidence theory to obtain the final classification results. The effectiveness and superiority of this method in processing small sample data are verified by experiments.

Intelligent prediction of rail corrugation evolution trend based on self-attention bidirectional TCN and GRU

Analyzing the evolution trend of rail corrugation using signal processing and deep learning is critical for railway safety, as current traditional methods struggle to capture the complex evolution of corrugation. This present study addresses the challenge of accurately capturing this trend, which relies significantly on expert judgment, by proposing an intelligent prediction method based on self-attention (SA), a bidirectional temporal convolutional network (TCN), and a bidirectional gated recurrent unit (GRU). First, multidomain feature extraction and adaptive feature screening were used to obtain the optimal feature set. These features were then combined with principal component analysis (PCA) and the Mahalanobis distance (MD) method to construct a comprehensive health indicator (CHI) that reflects the evolution of rail corrugation. A bidirectional fusion model architecture was employed to capture the temporal correlations between forward and backward information during corrugation evolution, with SA embedded in the model to enhance the focus on key information. The outcome was a rail corrugation trend prediction network that combined a bidirectional TCN, bidirectional GRU, and SA. Subsequently, a multi-strategy improved crested porcupine optimizer (CPO) algorithm was constructed to automatically obtain the optimal network hyperparameters. The proposed method was validated with on-site rail corrugation data, demonstrating superior predictive performance compared to other advanced methods. In summary, the proposed method can accurately predict the evolution trend of rail corrugation, offering a valuable tool for on-site railway maintenance.

Tool Wear Prediction Combining Global Feature Attention and Long Short-Term Memory Network

This study aims to accurately predict tool flank wear in milling and simplify the complexity of feature selection. A hybrid approach is proposed to eclectically integrate the advantages between the long short-term memory (LSTM) network and the global feature attention (GFA) module. First, the feature matrix is calculated using the multi-domain feature extraction method. Subsequently, a parallel network is employed to achieve feature fusion. The stacked LSTM network extracts the temporal dependencies between features and the GFA module is used to adaptively complement key features representing global information of samples. Finally, the output features are concatenated, and tool wear prediction is achieved through a fully connected layer. Experiments demonstrate the optimal performance in predicting tool flank wear. Specifically, using the proposed GFA-LSTM framework reduces the mean absolute error (MAE) by 36.9%, 17.7%, and 25.2% in three experiments compared to the simple LSTM, validating the effectiveness of the proposed method.

A comprehensive approach with DTW-driven IMF selection, multi-domain fusion, and TSA-based feature selection for compound fault diagnosis

Fault diagnosis in industrial machinery ensures operational efficiency and prevents downtimes. However, compound faults present a significant challenge due to their complex nature. This manuscript proposes a novel methodology for identifying compound faults, integrating advanced techniques to address the challenges present in existing methods. Firstly, the ICEEMDAN algorithm is employed for signal decomposition, enabling the extraction of noise-free IMFs crucial for identifying fault signatures. A DTW-based selection criteria is introduced to enhance the accuracy of IMF selection. The Bessel Transform is utilized for time–frequency conversion. HU invariant moments are incorporated as image features, and multi-domain feature fusion is implemented to understand compound faults comprehensively. A novel TSA-based feature selection method is proposed to select optimal features from the fused feature set. The methodology’s effectiveness is evaluated through a case study of compound gear-bearing faults. The results demonstrate the superiority of the approach, with a testing accuracy of 99 %.

Parallel Channel Feature Weighted Seizure Prediction Based on Multi-Scale Spatial and Temporal Factorization

Epileptic seizure prediction based on electroencephalography (EEG) plays an important role in the field. However, the existing epilepsy prediction methods have little modeling ability to capture the interaction between features, and the high redundancy of features leads to the limitations of model performance. In addition, the feature information guided by the multi-channel spatial location of the brain region is ignored. To solve these problems, this paper proposes a parallel channel feature-weighted seizure prediction network based on multi-scale temporal and spatial factorization (MS-STFM-PCFWNet). Specifically, the feature information of time domain and multi-channel spatial domain of brain region can be extracted by using feature matrix to fully learn the correlation between channels. Secondly, the multi-scale spatiotemporal Factorizer (MS-STFM) is utilized to combine and interact the features, and the correlation information between the features is captured. Finally, by combining the multi-scale Inception module with an efficient channel attention mechanism, a parallel channel feature weighted network (PCFWNet) is constructed to effectively learn multi-domain features and map the discriminant representation of epilepsy prediction. The proposed MS-STFM-PCFWNet is evaluated on public CHB-MIT and BONN datasets. The experimental results show that compared with the most advanced methods, the proposed method achieves excellent predictive performance, which can be used for early warning of epileptic seizures in specific patients.

Low-light image enhancement guided by multi-domain features for detail and texture enhancement

Low-light image enhancement holds significant value in the fields of computer vision and image processing, such as in applications like surveillance photography and medical imaging. Images captured in low-light environments typically suffer from significant noise levels, low contrast, and color distortion. Although existing low-light image enhancement techniques can improve image brightness and contrast to some extent, they often introduce noise or result in over-enhancement, leading to the loss of detail and texture. This paper introduces an innovative approach to low-light image enhancement by fusing spatial and frequency domain features while optimizing them with multiple loss functions. The core of the algorithm lies in its multi-branch feature extraction, multi-loss function constraints, and a carefully designed model structure. In particular, the model employs an encoder-decoder architecture, where the encoder extracts spatial features from the image, the Fourier feature extraction module captures frequency domain information, and the histogram feature encoder-decoder module processes global brightness distribution. These extracted features are then fused and reconstructed in the decoder to produce the enhanced image. In terms of loss functions, the algorithm combines perceptual loss, structural similarity loss, Fourier loss, and histogram loss to ensure comprehensive and natural enhancement effects. The novelty of this algorithm lies not only in its multi-branch feature extraction design but also in its unique model structure, which synergistically improves image quality across different domains, effectively preventing over-enhancement, and ultimately achieving a balanced enhancement of brightness, details, and texture. Experimental results on multiple datasets, including SIDD, LOL, MIT-Adobe-FiveK, and LOL-v2-synthetic, demonstrate that the proposed method outperforms existing techniques in terms of image detail, texture, and brightness. Specifically, it achieves a PSNR of 27.52 dB on the LOL dataset, surpassing Wavelet Diffusion by 1.19 dB. Additionally, on the LOL-v2-synthetic dataset, it achieves a PSNR of 29.56 dB, exceeding Wavelet Diffusion by 3.06 dB. These results demonstrate a significant enhancement in the visual quality of low-light images.

Overcoming the Barrier of Incompleteness: A Hyperspectral Image Classification Full Model.

Deep learning-based methods have shown promising outcomes in many fields. However, the performance gain is always limited to a large extent in classifying hyperspectral image (HSI). We discover that the reason behind this phenomenon lies in the incomplete classification of HSI, i.e., existing works only focus on a certain stage that contributes to the classification, while ignoring other equally or even more significant phases. To address the above issue, we creatively put forward three elements needed for complete classification: the extensive exploration of available features, adequate reuse of representative features, and differential fusion of multidomain features. To the best of our knowledge, these three elements are being established for the first time, providing a fresh perspective on designing HSI-tailored models. On this basis, an HSI classification full model (HSIC-FM) is proposed to overcome the barrier of incompleteness. Specifically, a recurrent transformer corresponding to Element 1 is presented to comprehensively extract short-term details and long-term semantics for local-to-global geographical representation. Afterward, a feature reuse strategy matching Element 2 is designed to sufficiently recycle valuable information aimed at refined classification using few annotations. Eventually, a discriminant optimization is formulized in accordance with Element 3 to distinctly integrate multidomain features for the purpose of constraining the contribution of different domains. Numerous experiments on four datasets at small-, medium-, and large-scale demonstrate that the proposed method outperforms the state-of-the-art (SOTA) methods, such as convolutional neural network (CNN)-, fully convolutional network (FCN)-, recurrent neural network (RNN)-, graph convolutional network (GCN)-, and transformer-based models (e.g., accuracy improvement of more than 9% with only five training samples per class). The code will be available soon at https://github.com/jqyang22/ HSIC-FM.

Efficient and novel multidomain feature analysis model with incremental optimizations for enhancing pediatric myelodysplastic syndrome detection

Pediatric myelodysplastic syndromes (MDS) are complicated, thus early and accurate diagnosis is essential for treatment planning and patient care. Diagnostic processes often use discrete data domain analysis, which reduces accuracy and delays diagnosis. This work addresses these limitations by introducing an advanced Multi domain Feature Analysis Model (MFAM) enhanced with incremental optimizations to improve pediatric MDS detection. Traditional pediatric MDS diagnosis relies on subjective evaluations and limited data fusion, not modern computational methods. These constraints may reduce diagnosis accuracy and postpone action. The proposed MFAM integrates data from Clinical History, Physical Examination, Blood Cell Counts, Peripheral Blood Smear, Bone Marrow Aspiration and Biopsy, Cytogenetic Analysis, Flow Cytometry, Genetic Testing, Iron Studies, and Bone Marrow Cytology to overcome these challenges. The MFAM increases feature variance by fusing Bidirectional Long Short-Term Memory (BiLSTM) with Bidirectional Gated Recurrent Units (BiGRU). Deep Q Learning with Graph Recurrent Convolutional Neural Networks (DQGRCNN) boosts efficiency. Additionally, the model integrates the Vector Autoregressive Moving Average with Exogenous Inputs (VARMAX) to facilitate early prediction of paediatric MDS. These enhancements have resulted in significant improvements in the precision of paediatric MDS detection by 4.5%, accuracy by 3.5%, recall by 2.3%, Area Under the Curve (AUC) by 1.5%, and specificity by 2.4% while reducing diagnostic delays by 8.5%. Furthermore, the model enhances the precision of predictive analysis by 2.9%, accuracy by 3.5%, recall by 2.5%, AUC by 2.9%, specificity by 5.5%, and reduces delays in predictive analysis by 8.5%. The MFAM presented in this paper revolutionizes the diagnosis and treatment of paediatric MDS by efficiently combining diverse diagnostic data, employing advanced transformation and fusion techniques, and optimizing responses through DQGRCNN. The integration of VARMAX further enables early prediction of the disease. MFAM will enhance diagnostic precision, therapy start, and clinical outcomes for young MDS patients.

Gait Speed Reserve in the general population-based 'Good Aging in Skåne' cohort study-distribution and associated factors.

Gait Speed Reserve (GSR) expresses a difference between fast and comfortable gait speed and may have an impact on everyday functioning. It was also hypothesized as a useful proxy measure of physiological reserve. However, height-normalizing values of GSR and its associated factors have not been evaluated in a general population of older adults. Therefore, we aimed to investigate the distribution of height-normalized GSR (HN-GSR) in an elderly population-based cohort from urban and rural areas (n = 4342) aged 60-93 years and evaluate associated physiological and lifestyle factors. Using linear mixed models, we identified gender and nine modifiable factors as significantly associated with HN-GSR across four age groups. Better handgrip strength, cognition and standing balance, higher physical activity level, larger calf circumference, and less smoking had positive associations with HN-GSR, while female gender, more leg pain, higher weight and, alcohol consumption had opposite effects. The Marginal R2 imply that this model explained 26% of the variance in HN-GSR. Physical activity and handgrip strength varied across age groups in impact on HN-GSR. The differences were however comparatively minor. In this large cohort study of older adults, we proposed for the first time that factors associated with HN-GSR represented multi-domain features that are in line with previous findings reported for GSR. Measuring HN-GSR/GSR may help clinicians identify early physiological impairments or unhealthy lifestyle habits, especially among older women, and may also have safety implications in daily life. Further work is needed to find out if measuring HN-GSR/GSR may be useful in identifying adverse health outcomes and overall physiological reserve.

Novel imbalanced fault diagnosis method based on generative adversarial networks with balancing serial CNN and Transformer (BCTGAN)

In industrial production engineering, machines cannot sustain faults for extended periods, resulting in limited fault data collection. This scarcity of data impedes the advancement of accurate fault diagnosis models, particularly in the context of small-sample, extremely imbalanced fault modeling, which poses a significant and challenging problem. In this paper, we introduce an innovative data augmentation approach for addressing the small-sample imbalanced problem: the balanced conditioning CNN (Convolutional Neural Networks)-Transformer architecture based Generative Adversarial Network (BCTGAN). Firstly, we employ a serial CNN and Transformer architecture as the foundation of the generator. Secondly, we introduce micro-enhancement and SMOTE (Synthetic Minority Oversampling Technique) enhancement of samples to facilitate model training. Micro-enhancement simply augments the sample size, while SMOTE enhancement enhances diversity. Together, these techniques balance the training process by introducing new loss function terms. Finally, outlier samples are mitigated through a designed multi-domain feature screening method. Despite the added time cost and computational complexity, results from three instances demonstrate that the proposed method holds superior diagnostic results with great potential compared to traditional and general GAN-based fault diagnosis methods. Our work provides theoretical and practical insights for modeling the field of small-sample blended disequilibrium fault diagnosis.

Bearing fault diagnosis method based on multi-domain feature fusion and heterogeneous network under small sample conditions

Bearing fault diagnosis method based on multi-domain feature fusion and heterogeneous network under small sample conditions

Deep learning fusion framework for automated coronary artery disease detection using raw heart sound signals

One of the most common cardiovascular diseases is coronary artery disease (CAD). Thus, it is crucial for early CAD diagnosis to control disease progression. Computer-aided CAD detection often converts heart sounds into graphics for analysis. However, this method relies heavily on the subjective experience of experts. Therefore, in this study, we proposed a method for CAD detection using raw heart sound signals by constructing a fusion framework with two CAD detection models: a multidomain feature model and a medical multidomain feature fusion model. We collected heart sound signal datasets from 400 participants, extracting 206 multidomain features and 126 medical multidomain features. The designed framework fused the same one-dimensional deep learning features with different multidomain features for CAD detection. The experimental results showed that the multidomain feature model and the medical multidomain feature fusion model achieved areas under the curve (AUC) of 94.7 % and 92.7 %, respectively, demonstrating the effectiveness of the fusion framework in integrating one-dimensional and cross-domain heart sound features through deep learning algorithms, providing an effective solution for noninvasive CAD detection.

Rapid Detection Technology for Cocaine and Methamphetamine Acute Abuse based on EEG

Abstract Multi-domain characterization and a new fast detection method for acute illegal psychostimulants abuse detection based on waking-state EEG of mice is proposed in this paper. To get corftical electroencephalogram (EEG), three groups of mice were injected with cocaine (Coca), methamphetamine (Meth), and saline (Sali) respectively following the experimental paradigm of drug abuse. The EEG features were extracted out by multi-domain views, including time, frequency, complexity, dynamics, and independent domains to obtain acute drug abuse effects on the brain. New strategy combing time domain with frequency domain is developed as multi-domain input and by means of dimension transformation approach preserving temporal information, the performance of drug abuse detection is greatly improved with deep learning models of ResNet50. Results show that comparing to support vector machine (SVM), k nearest neighbor (kNN), random forest (RF), and long short-term memory fully convolutional networks (LSTM-FCN), Res-Net50 based on our proposed multi-domain features has best F1-score of 95.10%. This promising method provides a low-cost, fast, and widely assisted technology for psychostimulants abuse identification.

Remaining useful life prediction for a cracked rotor system via moving feature fusion based deep learning approach

This study proposes a novel approach for predicting the remaining useful life (RUL) of a cracked rotor system by using a moving fusion gated recurrent unit (MFGRU) model. Firstly, 12 multi-domain features were extracted from the raw collected vibration signals, which were then fused and distributed adaptively using bidirectional exponential moving average (EMA) and multi-head attention (MHA). Then a bidirectional gated recurrent unit network combined with moving feature fusion method was proposed to capture the long-term dependence relationship in the monitoring data of the cracked rotor for RUL prediction. Finally, the model’s performance was validated through accelerated life experiments, with the measured values of root mean square error (RMSE) and mean absolute error (MAE) below 3.17 and 2.60, respectively. This study offers a dynamic RUL prediction method with certain significance and valuable reference for designing deep learning models.

Radio Signal Modulation Recognition Method Based on Hybrid Feature and Ensemble Learning: For Radar and Jamming Signals.

The detection performance of radar is significantly impaired by active jamming and mutual interference from other radars. This paper proposes a radio signal modulation recognition method to accurately recognize these signals, which helps in the jamming cancellation decisions. Based on the ensemble learning stacking algorithm improved by meta-feature enhancement, the proposed method adopts random forests, K-nearest neighbors, and Gaussian naive Bayes as the base-learners, with logistic regression serving as the meta-learner. It takes the multi-domain features of signals as input, which include time-domain features including fuzzy entropy, slope entropy, and Hjorth parameters; frequency-domain features, including spectral entropy; and fractal-domain features, including fractal dimension. The simulation experiment, including seven common signal types of radar and active jamming, was performed for the effectiveness validation and performance evaluation. Results proved the proposed method's performance superiority to other classification methods, as well as its ability to meet the requirements of low signal-to-noise ratio and few-shot learning.

Dynamic Data-Driven degradation method for monitoring remaining useful life of cutting tools

Monitoring tool condition and remaining useful life (RUL) are vital in preventing the occurrence of excessive tool wear. This paper develops a novel dynamic data-driven degradation method for monitoring the RUL of cutting tools. In sensor-data collection, vibration, sound, and power external sensors and built-in data are gathered from the machine tool. In multi-feature selection and dynamic model updating, a decision-level fusion method of spanning multi-domain features is designed to dynamically utilize a global prediction error for selecting and fusion degradation features, which are associated with the cutting tool life in the degradation model. A rolling HI-RUL mapping is established in RUL prediction by employing a historical health indicator curve, which estimates the RUL of cutting tool with a given threshold. The effectiveness of proposed method was assessed through two run-to-failure experiments of cutting tools, showing that an average global prediction error is reduced to 4.07%.

Fused feature extract method for Φ-OTDR event recognition based on VGGish transfer learning

Thanks to the development of artificial intelligence algorithms, the event recognition of distributed optical fiber sensing systems has achieved high classification accuracy on many deep learning models. However, the large-scale samples required for the deep learning networks are difficult to collect for the optical fiber vibration sensing systems in actual scenarios. An overfitting problem due to insufficient data in the network training process will reduce the classification accuracy. In this paper, we propose a fused feature extract method suitable for the small dataset of Φ-OTDR systems. The high-dimensional features of signals in the frequency domain are extracted by a transfer learning method based on the VGGish framework. Combined with the characteristics of 12 different acquisition points in the space, the spatial distribution characteristics of the signal can be reflected. Fused with the spatial and temporal features, the features undergo a sample feature correction algorithm and are used in a SVM classifier for event recognition. Experimental results show that the VGGish, a pre-trained convolutional network for audio classification, can extract the knowledge features of Φ-OTDR vibration signals more efficiently. The recognition accuracy of six types of intrusion events can reach 95.0% through the corrected multi-domain features when only 960 samples are used as the training set. The accuracy is 17.7% higher than that of the single channel trained on VGGish without fine-tuning. Compared to other CNNs, such as ResNet, the feature extract method proposed can improve the accuracy by at least 4.9% on the same dataset.

Prediction method of human defecation based on informer audio data augmentation and improved residual network

Defecation care for disabled patients is a major challenge in health management. Traditional post-defecation treatment will bring physical injury and negative emotions to patients, while existing pre-defecation forecasting care methods are physically intrusive. On the basis of exploring the mechanism of defecation intention generation, and based on the characteristic analysis and clinical application of bowel sounds, it is found that the generation of desire to defecate and bowel sounds are correlated to a certain extent. Therefore, a deep learning-based bowel sound recognition method is proposed for human defecation prediction. The wavelet domain based Wiener filter is used to filter the bowel sound data to reduce other noise. Statistical analysis, fast Fourier transform and wavelet packet transform are used to extract the integrated features of bowel sound in time, frequency and time-frequency domain. In particular, an audio signal expansion data algorithm based on the Informer model is proposed to solve the problem of poor generalization of the training model caused by the difficulty of collecting bowel sound in reality. An improved one-dimensional residual network model (1D-IResNet) for defecation classification prediction is designed based on multi-domain features. The experimental results show that the proposed bowel sound augmentation strategy can effectively improve the data sample size and increase the sample diversity. Under the augmented dataset, the training speed of the 1D-IResNet model is accelerated, and the classification accuracy reaches 90.54 %, the F1 score reaches 83.88 %, which achieves a relatively good classification stability while maintaining a high classification index.