Feature Extraction Classification Research Articles

Attribute disentanglement and re-entanglement for generalized zero-shot learning

The key challenge in zero-shot learning is inferring latent semantic knowledge between visual and attribute features of seen classes to achieve knowledge transfer to unseen classes. To address the limitation that local attribute features can only ensure attribute-level recognition rather than classification of an entire class, some methods incorporate global information into the process or results of local features extraction for classification. However, these approaches have not effectively addressed the issue. To address these issues, we propose an Attribute Disentanglement and Re-entanglement for Generalized Zero-Shot Learning. Our model no longer implicitly or explicitly incorporates global information into local attribute features for classification. Instead, we adjust local attribute features to make them more suitable for classification in re-entanglement phase, while ensuring the correct extraction of these features in disentanglement phase. We employ appropriate optimization loss functions and achieve significant improvements on three challenging benchmark datasets. Compared to other similar methods, our model exhibits strong competitiveness.

An Ensemble Machine Learning Approach for Sea Ice Monitoring Using CFOSAT/SCAT Data

Sea ice is a crucial component of the global climate system. The China–French Ocean Satellite Scatterometer (CFOSAT/SCAT, CSCAT) employs an innovative rotating fan beam system. This study applied principal component analysis (PCA) to extract classification features and developed an ensemble machine learning approach for sea ice detection. PCA identified key features from CSCAT’s backscatter information, representing outer and sweet swath observations. The ensemble model’s performances (OA and Kappa) for the Northern and Southern Hemispheres were 0.930, 0.899, and 0.844, 0.747, respectively. CSCAT achieved an accuracy of over 0.9 for close ice and open water but less than 0.3 for open ice, with misclassification of open ice as closed ice. The sea ice extent discrepancy between CSCAT and the National Snow and Ice Data Center (NSIDC) was −0.06 ± 0.36 million km2 in the Northern Hemisphere and −0.03 ± 0.48 million km2 in the Southern Hemisphere. CSCAT’s sea ice closely matched synthetic aperture radar (SAR) imagery, indicating effective sea ice and open water differentiation. CSCAT accurately distinguished sea ice from open water but struggled with open ice classification, with misclassifications in the Arctic’s Greenland Sea and Hudson Bay, and the Antarctic’s sea ice–water boundary.

Exploring the potential of representation and transfer learning for anatomical neuroimaging: Application to psychiatry

The perspective of personalized medicine for brain disorders requires efficient learning models for anatomical neuroimaging-based prediction of clinical conditions. There is now a consensus on the benefit of deep learning (DL) in addressing many medical imaging tasks, such as image segmentation. However, for single-subject prediction problems, recent studies yielded contradictory results when comparing DL with Standard Machine Learning (SML) on top of classical feature extraction. Most existing comparative studies were limited in predicting phenotypes of little clinical interest, such as sex and age, and using a single dataset. Moreover, they conducted a limited analysis of the employed image pre-processing and feature selection strategies. This paper extensively compares DL and SML prediction capacity on five multi-site problems, including three increasingly complex clinical applications in psychiatry namely schizophrenia, bipolar disorder, and Autism Spectrum Disorder (ASD) diagnosis. To compensate for the relative scarcity of neuroimaging data on these clinical datasets, we also evaluate three pre-training strategies for transfer learning from brain imaging of the general healthy population: self-supervised learning, generative modeling and supervised learning with age. Overall, we find similar performance between randomly initialized DL and SML for the three clinical tasks and a similar scaling trend for sex prediction. This was replicated on an external dataset. We also show highly correlated discriminative brain regions between DL and linear ML models in all problems. Nonetheless, we demonstrate that self-supervised pre-training on large-scale healthy population imaging datasets (N≈10k), along with Deep Ensemble, allows DL to learn robust and transferable representations to smaller-scale clinical datasets (N≤1k). It largely outperforms SML on 2 out of 3 clinical tasks both in internal and external test sets. These findings suggest that the improvement of DL over SML in anatomical neuroimaging mainly comes from its capacity to learn meaningful and useful abstract representations of the brain anatomy, and it sheds light on the potential of transfer learning for personalized medicine in psychiatry

Semi-Supervised Autoencoder for Chemical Gas Classification with FTIR Spectrum.

Chemical warfare agents pose a serious threat due to their extreme toxicity, necessitating swift the identification of chemical gases and individual responses to the identified threats. Fourier transform infrared (FTIR) spectroscopy offers a method for remote material analysis, particularly in detecting colorless and odorless chemical agents. In this paper, we propose a deep neural network utilizing a semi-supervised autoencoder (SSAE) for the classification of chemical gases based on FTIR spectra. In contrast to traditional methods, the SSAE concurrently trains an autoencoder and a classifier attached to a latent vector of the autoencoder, enhancing feature extraction for classification. The SSAE was evaluated on laboratory-collected FTIR spectra, demonstrating a superior classification performance compared to existing methods. The efficacy of the SSAE lies in its ability to generate denser cluster distributions in latent vectors, thereby enhancing gas classification. This study established a consistent experimental environment for hyperparameter optimization, offering valuable insights into the influence of latent vectors on classification performance.

Retraction Note: Hermite-based texture feature extraction for classification of humeral head in proton density-weighted MR images

Retraction Note: Hermite-based texture feature extraction for classification of humeral head in proton density-weighted MR images

Enhanced X-wave recognition in ECG signal using max–min thresholds and detection of QRS complex

Due to the advantages of Electrocardiogram (ECG) signals, which are challenging to replicate yet easy to get, ECG-based identification has become a new path in biometric recognition research. These classic feature extraction techniques require Hand-crafted or feature-specific implications. The methods used for selection and integration of features, are time-consuming. The main objective of this study is develop deep learning approach to study the features of ECG data digital characteristics, thus saving a lot of signal pre-processing steps. This research proposed novel technique in X-wave recognition of ECG signal using max-min threshold technique and classification of ECG signal. This signal has been processed for noise removal and normalization. Then this processed signal has been used to recognize X-wave from ECG signal. From recognized X-wave, the ECG signal has been classified using Improved Support Vector Machine (ISVM). The QRS complex has been detected using Stacked Auto-Encoder with Neural Networks (SAENN). The study took raw ECG signals and entropy-based features evaluated from extracted QRS complexes. Exams are based on classifying heart disorders into two, five, and twenty classes. The experimental findings showed that our suggested model attained a high classification accuracy of 97%, precision of 89%, recall of 90%, F-1 score of 88%.

A Learnable Counter-Condition Analysis Framework for Functional Connectivity-Based Neurological Disorder Diagnosis.

To understand the biological characteristics of neurological disorders with functional connectivity (FC), recent studies have widely utilized deep learning-based models to identify the disease and conducted post-hoc analyses via explainable models to discover disease-related biomarkers. Most existing frameworks consist of three stages, namely, feature selection, feature extraction for classification, and analysis, where each stage is implemented separately. However, if the results at each stage lack reliability, it can cause misdiagnosis and incorrect analysis in afterward stages. In this study, we propose a novel unified framework that systemically integrates diagnoses (i.e., feature selection and feature extraction) and explanations. Notably, we devised an adaptive attention network as a feature selection approach to identify individual-specific disease-related connections. We also propose a functional network relational encoder that summarizes the global topological properties of FC by learning the inter-network relations without pre-defined edges between functional networks. Last but not least, our framework provides a novel explanatory power for neuroscientific interpretation, also termed counter-condition analysis. We simulated the FC that reverses the diagnostic information (i.e., counter-condition FC): converting a normal brain to be abnormal and vice versa. We validated the effectiveness of our framework by using two large resting-state functional magnetic resonance imaging (fMRI) datasets, Autism Brain Imaging Data Exchange (ABIDE) and REST-meta-MDD, and demonstrated that our framework outperforms other competing methods for disease identification. Furthermore, we analyzed the disease-related neurological patterns based on counter-condition analysis.

Sparse Trace Ratio LDA for Supervised Feature Selection.

Classification is a fundamental task in the field of data mining. Unfortunately, high-dimensional data often degrade the performance of classification. To solve this problem, dimensionality reduction is usually adopted as an essential preprocessing technique, which can be divided into feature extraction and feature selection. Due to the ability to obtain category discrimination, linear discriminant analysis (LDA) is recognized as a classic feature extraction method for classification. Compared with feature extraction, feature selection has plenty of advantages in many applications. If we can integrate the discrimination of LDA and the advantages of feature selection, it is bound to play an important role in the classification of high-dimensional data. Motivated by the idea, we propose a supervised feature selection method for classification. It combines trace ratio LDA with l2,p -norm regularization and imposes the orthogonal constraint on the projection matrix. The learned row-sparse projection matrix can be used to select discriminative features. Then, we present an optimization algorithm to solve the proposed method. Finally, the extensive experiments on both synthetic and real-world datasets indicate the effectiveness of the proposed method.

Slow Down to Go Better: A Survey on Slow Feature Analysis.

Temporal data contain a wealth of valuable information, playing an essential role in various machine-learning tasks. Slow feature analysis (SFA), one of the most classic temporal feature extraction models, has been deeply explored in two decades of development. SFA extracts slowly varying features as high-level representations of temporal data. Its core idea of "slow" has been proven to be consistent with the nature of biological vision and beneficial in capturing significant temporal information for various tasks. So far, SFA has evolved into numerous improved versions and is widely applied in many fields such as computer vision, industrial control, remote sensing, signal processing, and computational biology. However, there currently lacks an insightful review of SFA. In this article, a comprehensive overview of SFA and its extensions is provided for the first time. The formulation and optimization of SFA are introduced. Two mainstream solutions, geometric interpretation, and a gradient-based training method of SFA are presented and discussed. Following that, a taxonomy of the current progress of SFA is proposed. We classify improved versions of SFA into six categories, including dual-input SFA (DISFA), online slow feature analysis (OSFA), probabilistic SFA (PSFA), multimode SFA, nonlinear SFA, and discrete labeled SFA. For each category, we illustrate its main ideas, mathematical principles, and applicable scenarios. In addition, the practical applications of SFA are summarized and presented. Finally, we bring new insights into SFA according to its research status and provide potential research directions, which may serve as a good reference for promoting future work.

Classification of self-limited epilepsy with centrotemporal spikes by classical machine learning and deep learning based on electroencephalogram data

Electroencephalogram (EEG) has been widely utilized as a valuable assessment tool for diagnosing epilepsy in hospital settings. However, clinical diagnosis of patients with self-limited epilepsy with centrotemporal spikes (SeLECTS) is challenging due to the presence of similar abnormal discharges in EEG displays compared to other types of epilepsy (non-SeLECTS) patients. To assist the diagnostic process of epilepsy, a comprehensive classification study utilizing machine learning or deep learning techniques is proposed. In this study, clinical EEG was collected from 33 patients diagnosed with either SeLECTS or non-SeLECTS, aged between 3 and 11 years. In the realm of classical machine learning, sharp wave features (including upslope, downslope, and width at half maximum) were extracted from the EEG data. These features were then combined with the random forest (RF) and extreme random forest (ERF) classifiers to differentiate between SeLECTS and non-SeLECTS. Additionally, deep learning was employed by directly inputting the EEG data into a deep residual network (ResNet) for classification. The classification results were evaluated based on accuracy, F1-score, area under the curve (AUC), and area under the precision-recall curve (AUPRC). Following a 10-fold cross-validation, the ERF classifier achieved an accuracy of 73.15 % when utilizing sharp wave feature extraction for classification. The F1-score obtained was 0.72, while the AUC and AUPRC values were 0.75 and 0.63, respectively. On the other hand, the ResNet model achieved a classification accuracy of 90.49 %, with an F1-score of 0.90. The AUC and AUPRC values for ResNet were found to be 0.96 and 0.92, respectively. These results highlighted the significant potential of deep learning methods in SeLECTS classification research, owing to their high accuracy. Moreover, feature extraction-based methods demonstrated good reliability and could assist in identifying relevant biological features of SeLECTS within EEG data.

Design of innovative teaching methods for classical dance based on group intelligence

Abstract Compared with image records, 3D dance scores can realize the demonstration and practice of classical dance teaching more accurately and efficiently. In this paper, the Li group skeletal representation model is used for Li algebra space mapping processing to learn action feature data, classify them, and complete classical dance action feature extraction. Then, the residual network model is used as the backbone network to construct a two-stream spatiotemporal residual network for action recognition by using two different temporal spans and fusing the results of the two-channel networks through side connections. Finally, the group intelligence algorithm group intelligence algorithm ACOR-SGD is designed to optimize the parameters of the convolutional neural network, and then the dance score generation model is integrated into the design of the innovative teaching method of classical dance for the application of the professional teaching of classical dance in colleges and universities. There is basically no significant difference between the pre-test scores of the experimental and control classes. After one semester of applying the innovative teaching method based on the generation of dance scores, the scores of the experimental class in the five items of basic pace mastery, movement fluency, stage presence, technical quality, and coordination of the music rhythm were improved by 4.6, 4, 4.28, 3.63, and 4.72 respectively, and the total score of 81.54 was much higher than that of the control class of 60.57 points posttest scores, and all were significant at the 0.01 level. This study explores innovatively and provides an effective path for integrating dance teaching with cutting-edge information technology, thus improving the quality and efficiency of dance teaching.

A Study of the Aesthetic Application of Machine Learning to Five Education in Classical Music Interpretation

Abstract With the application of machine learning to the teaching of classical music in the context of five education as the research objective, this study first tested the feature extraction performance and classification of classical music with specific classical music data. Secondly, it explored the changes in the three elements of song aesthetics before and after machine learning five-education music teaching in two academic periods and analyzed the teaching effect it had on the task completion time and rhythmic accuracy of the songs. The use of audio analysis software in terms of task completion time, rhythmic accuracy, and pitch accuracy dimensions produces quantitative teaching effect ratings. The results show that the method in this paper obtains ideal classical music classification and feature extraction results and the ratings of the music feature extraction algorithms are in line with the real level of teachers and students. The experimental group and the control group, after teaching music using machine learning, took a similar time to learn the task, but the difference in rhythmic accuracy was large. The highest IOI value of the experimental group is only 0.07439, and the overall average of the group’s satisfaction with the teaching effect has increased by 0.842. This paper’s method is conducive to the development and innovation of music teaching.

Fault Classification in Reciprocating Compressors: A Comparison of Machine Learning and Deep Learning Approaches

This study compares methodologies for fault classification in reciprocating compressors, focusing on traditional Machine Learning (ML) with classical feature extraction processes and one-dimensional Convolutional Neural Networks (1D-CNN) in Deep Learning (DL). Both techniques demonstrated viability by employing a dataset of compressor vibration signals encompassing ten fault classes. While ML achieved a classification accuracy of 86%, DL reached 90.709%, highlighting its superior learning and generalization abilities, although with longer training times. These findings suggest that, despite ML being effective when relevant prior knowledge is available, DL, particularly with 1D-CNN, offers enhanced fault classification performance for this study case at the expense of additional processing resources.

Classification of Congestive Heart Failure Using Artificial Neural Network Based on Higher-Order Moments Detrended Fluctuation Analysis of Heart Rate Variability

Coronary heart disease or cardiovascular disease is a kind of heart disease that affects the heart and all the blood vessels in the body caused by a build-up of plaque in a person's arteries and can result in a stroke or heart attack. According to the Indonesian Ministry of Health, (Depkes RI) coronary heart disease is the main and first cause of all deaths by heart disease, which account for 26.4% among all other causes. Considering the severity of the condition, it is necessary to detect this kind of heart disease to reduce the number of deaths from heart disease. Heart disease detection can be performed by checking the Heart Rate Variability (HRV) signal. HRV assessment is carried out by analyzing short-term and long-term Electrocardiogram (ECG) records. To analyze HRV signals, a Higher-Order Moments Detrended Fluctuation Analysis feature extraction method is proposed to see the non-linear structure of HRV data. The input for Higher-Order Moments Detrended Fluctuation Analysis is an ECG signal which has been converted into an HRV signal. The output of the feature extraction shows that the value of kurtosis-based fluctuation function has a statistical significance in order to differentiate the HRV of heart disease patients and that of normal subjects. These values are then used as the input for classification using an Artificial Neural Network. The output of the classification is the classification of patients with heart disease and normal subjects. The results of the Higher-Order Moments Detrended Fluctuation Analysis feature extraction classification yield the best accuracy of 71.43% with a ROC value of 0.774 which can be categorized as a pretty good classification.

A Time–Frequency Residual Convolution Neural Network for the Fault Diagnosis of Rolling Bearings

A time–frequency residual convolution neural network (TFRCNN) was proposed to identify various rolling bearing fault types more efficiently. Three novel points about TFRCNN are presented as follows: First, by constructing a double-branch convolution network in the time domain and the frequency domain, the respective features in the time domain and the frequency domain were extracted to ensure the rich and complete feature representation of raw data sources. Second, specific residual structures were designed to prevent learning degradation of the deep network, and global average pooling was adopted to improve the network’s sparsity. Third, TFRCNN was better than the other models in terms of prediction accuracy, robustness, generalization ability, and convergence. The experimental results demonstrate that the prediction accuracy rate of TFRCNN, trained using mixing load data, reached 98.88 to 99.92% after optimizing the initial learning rate and choosing the optimizer and loss function. It was verified that TFRCNN can adaptively learn to extract deep fault features, accurately identify bearing fault conditions, and overcome the limitations of classical shallow feature extraction and classification methods, as well as common convolution neural networks. Hence, this investigation revealed TFRCNN’s potential for bearing fault diagnosis in practical engineering applications.

Application of multi-source data fusion of modern dress design with the artistic elements of Chinese embroidery

Abstract This paper first investigates the method of extracting features from Chinese embroidery art images, including preprocessing, filtering, and enhancement steps. Then, the multi-classification recognition technique based on an improved support vector machine is used to classify and recognize different modern dress designs. Non-iterative and conjugate orthogonal algorithms are proposed for a feature fusion method based on partial least squares analysis. In the performance analysis part, the performance of feature extraction classification, data fusion energy, and fused data noise are evaluated. Lastly, the impact of the dress style generation map incorporating fused Chinese embroidery elements is analyzed. The study shows that the method in this paper can save 33.4 J of energy consumption after running for 250 s compared with the popular learning-based method. At least a 10% improvement in the number of dress style graphs that interest social event participants, etiquette service practitioners, and performers was achieved after multiple training iterations.

A Machine Learning Approach for Simultaneous Classification of Material Types and Cracks

Exterior structures are susceptible to deformation, which can manifest as cracks on the surface. Deformations that occur on surfaces subjected to daily human use can exacerbate rapidly, potentially leading to irreversible structural damage. They have a potential to result in fatalities. Thus, continuous inspection of these deformations is of invaluable importance. In addition, the identification of the materials comprising the structures is essential to facilitate the implementation of appropriate precautionary measures. However, the inspections are hard to maintain with a solely human workforce. More advanced actions can be taken thanks to the developments in technology. Machine Learning methods could be used in this area where human workforce is ineffective. In this regard, an end-to-end Machine Learning approach was proposed in this study. The power of classical feature extraction methods and Artificial Neural Networks were combined to detect cracks and material of the surface simultaneously. The 2D Discrete Wavelet Transform and statistical properties gained from Gray Level Co-Occurrence Matrix were utilized in the feature extraction mechanism, and an ANN structure was designed. The findings of the study indicate that the proposed mechanism achieved an acceptable level of accuracy for recognizing the structural deformations, despite the challenges posed by the complexity of the problem.

Hybrid framework for respiratory lung diseases detection based on classical CNN and quantum classifiers from chest X-rays

The human respiratory system might be seriously affected by COVID-19 infection. Therefore, early classification of it is a crucial task. Quantum machine learning and quantum neural network models can play an effective role in multiclass classification problems. Compared to standard deep and machine learning classifiers, the quantum variational classifier may lead to less memory usage, accuracy, and portability for respiratory disease detection. This article proposes a hybrid respiratory lung disease detection framework based on classical CNN and Quantum classifiers. It combines a classical deep feature extraction model with quantum classifiers. A new custom convolutional neural network (CCNN) deep learning model is proposed to perform feature extraction, and the Multi-Multi-Single (MMS) & Multi-Single-Multi-Single (MSMS) are proposed as quantum machine learning algorithms. These two quantum classifiers are built via a quantum variational circuit with encoding, entanglement, and measurement properties. The tests were carried out on the COVID-19 Radiography Dataset (CRD), which contains 15,153 chest X-ray images of COVID-19, Viral, and Normal. The experimental results revealed that the proposed model had the highest training and testing accuracy of 98.9% and 98.1%, on the CRD dataset, with a computation cost of 0.07 and 0.08 respectively. This hybrid model performs better than the other standard deep learning models. Additionally, we validated our MMS and MSMS quantum classifiers by deploying them on the IBM Q-QASM real-time quantum computer.

Liver Lesion Detection Using Semantic Segmentation and Chaotic Cuckoo Search Algorithm

The classic feature extraction techniques used in recent research on computer-aided diagnosis (CAD) of liver cancer have several disadvantages, including duplicated features and substantial computational expenses. Modern deep learning methods solve these issues by implicitly detecting complex structures in massive quantities of healthcare image data. This study suggests a unique bio-inspired deep-learning way for improving liver cancer prediction outcomes. Initially, a novel semantic segmentation technique known as UNet++ is proposed to extract liver lesions from computed tomography (CT) images. Second, a hybrid approach that combines the Chaotic Cuckoo Search algorithm and AlexNet is indicated as a feature extractor and classifier for liver lesions. LiTS, a freely accessible database that contains abdominal CT images, was employed for liver tumor diagnosis and investigation. The segmentation results were evaluated using the Dice similarity coefficient and Correlation coefficient. The classification results were assessed using Accuracy, Precision, Recall, F1 Score, and Specificity. Concerning the performance metrics such as accuracy, precision, and recall, the recommended method performs better than existing algorithms producing the highest values such as 99.2%, 98.6%, and 98.8%, respectively.

MTDN: Learning Multiple Temporal Dynamics Representation for Emotional Valence Classification with EEG.

Emotion recognition from electroencephalogram (EEG) requires computational models to capture the crucial features of the emotional response to external stimulation. Spatial, spectral, and temporal information are relevant features for emotion recognition. However, learning temporal dynamics is a challenging task, and there is a lack of efficient approaches to capture such information. In this work, we present a deep learning framework called MTDN that is designed to capture spectral features with a filterbank module and to learn spatial features with a spatial convolution block. Multiple temporal dynamics are jointly learned with parallel long short-term memory (LSTM) embedding and self-attention modules. The LSTM module is used to embed the time segments, and then the self-attention is utilized to learn the temporal dynamics by intercorrelating every embedded time segment. Multiple temporal dynamics representations are then aggregated to form the final extracted features for classification. We experiment on a publicly available dataset, DEAP, to evaluate the performance of our proposed framework and compare MTDN with existing published results. The results demonstrate improvement over the current state-of-the-art methods on the valence dimension of the DEAP dataset.