Feature Extraction Method Research Articles

Underwater acoustic target recognition based on multi-resolution wavelet feature deep learning

The application of underwater acoustic target recognition is extensive, exploration of marine resources, submarine vehicle detection and so on. However, traditional feature extraction methods are susceptible to the influence of complex marine environments and target operating conditions, leading to a significant reduction in the correct recognition rate. This paper introduces an underwater acoustic target recognition method that integrates multi-resolution wavelet features with deep learning algorithms, aiming to overcome the limitations of traditional time-frequency analysis techniques, which are unable to extract multiple signal characteristics simultaneously due to the trade-off between time and frequency resolution. Specifically, the essence of this technique is twofold: (1) Selecting suitable wavelet bases and decomposition levels to capture signal characteristics at different resolutions, effectively seizing the signal's local features within the time-frequency domain; (2) Fusing signal features across different resolutions to optimize the feature extraction process and enhance the distinguishability of target features. Finally, by applying deep learning algorithms to experimentally measured underwater acoustic data, the results demonstrate that this method can effectively enhance the accuracy of underwater acoustic target recognition.

Research on Rotating Machinery Fault Diagnosis Based on Multi-Strategy Feature Extraction

This study proposes a novel feature extraction method that leverages multi-strategy optimization algorithms to enhance the accuracy and efficiency of rotating machinery fault diagnosis. By introducing an improved slime mold algorithm (LTSSMA), it optimizes the penalty factor and decomposition layers in variational mode decomposition (VMD), achieving more precise signal decomposition. The sample entropy generated by VMD forms the basis of the feature vector, and the nonlinear dimensionality reduction algorithm (t-SNE) is applied to reduce dimensions. Finally, the optimized features are classified using a random forest (RF) model, resulting in an 11.3% improvement in fault diagnosis accuracy compared to traditional methods. This method not only accelerates the diagnostic process but also significantly improves fault identification reliability.

Application of SPNGO-VMD-SVM in rolling bearing fault diagnosis

Abstract Traditional bearing fault feature extraction and fault classification methods have low recognition accuracy and limited recognition capability in noisy environments. To address this problem, this paper proposes an improved northern eagle algorithm (SPNGO) to optimize the variational modal decomposition (VMD) and support vector machine (SVM) to achieve bearing fault diagnosis. Firstly, to overcome the disadvantages of NGO, such as easy fall into local optimal solutions and slow convergence speed, the Sine Cosine Strategy (SCA) and Position Optimisation Search Algorithm (POS) are introduced to optimize the NGO, and the superiority of the SPNGO algorithm is proved through the comparison of different algorithms. Then, SPNGO-VMD is used to adaptively decompose the vibration signals of faulty bearings and generate multiple modal components IMF. The effective IMF components are screened based on the craggy principle to reconstruct the signals. Finally, the reconstructed feature signals are input into SPNGO-SVM for fault classification and compared with other fault diagnosis models. The study results show that the fault diagnosis accuracy of SPNGO-VMD-SVM can reach 96.67%, which can effectively improve the accuracy of bearing fault diagnosis.

Comparative analysis of feature selection and extraction methods for student performance prediction across different machine learning models

Education is at the core of developmental progress, necessitating the exploration and implementation of diverse contemporary methods to ensure the success of students across multiple levels. However, impediments to this success exist, categorized into three primary groups which are, individual factors, family factors and social factors. These factors can manifest as absenteeism and boredom, posing a threat to the future of both students and society at large. Addressing these challenges proves to be a complex task for educators and pedagogues, given the unique problems each student faces. This paper aims to employ a broad spectrum of feature selection and extraction methods, some unconventional in the education sector but proven reliable in other domains. By integrating machine learning (ML) and deep learning (DL) models, we seek to predict student performance based on these identified factors. Subsequently, a comparative analysis will be conducted to determine the most effective model, considering the relevance of various factors.

A Novel Fault Feature Selection and Diagnosis Method for Rotating Machinery with SI-IR2CMSE and SSGMM-SR

Abstract In the field of fault diagnosis, machine learning is highly valued for its broad applicability and efficiency. Feature extraction and feature selection are key steps in the application of machine learning, and the performance of fault diagnosis methods relies heavily on the effective execution of these two steps. For this reason, this paper aims to enhance the performance of fault diagnosis methods by improving these two aspects. Firstly, to address the non-linearity and non-stationarity of rotating machinery vibration signals under variable operation conditions, this paper proposes an improved rapid refined composite multiscale sample entropy (IR2CMSE) feature extraction method. In addition, this paper decomposes the vibration signals with improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) and extracts the sensitive intrinsic modal functions’ (IMFs') IR2CMSE values (SI-IR2CMSE) as the initial feature vector, which more accurately reveals the intrinsic time-scale characteristics of the vibration signals. Secondly, to address the problem of over-reliance on sample labels in most feature selection methods, this paper proposes a semi-supervised Gaussian mixing model with sparse regularization (SSGMM-SR) feature selection model. The model does not require complete fault labels and can automatically identify important features. Finally, validation with two rotating machinery fault datasets shows that the method proposed in this study exhibits high diagnostic accuracy and stability across multiple classifiers.

Knowledge-informed FIR-based cross-category filtering framework for interpretable machinery fault diagnosis under small samples

Relying on sufficient training data, the existing fault diagnosis methods rarely focus on the methodological interpretability and the data scarcity in real industrial scenarios simultaneously. Motivated by this issue, we deeply reexamined the intrinsic characteristics of fault signals and the guiding significance of classical signal-processing methods for feature enhancement. From the perspective of multiscale modes, this study tailors multiple learnable knowledge-informed finite impulse response (FIR) filtering kernels to extract sensitive modes for explainable feature enhancement. On this foundation, a knowledge-informed FIR-based cross-category filtering (FIR-CCF) framework is further proposed for interpretable small-sample fault diagnosis. With the consideration of the mode complexity, a cross-category filtering strategy is explored to further enhance feature expressions for identifying single state. To be special, this strategy divides a multi-class recognition process into multiple two-class recognition task. A multi-task learning is then presented where multiple binary-class base learners (BCBLearners) that consists of a feature extractor and a two-class classifier is established to seek discriminate mode features for each type of state. Eventually, all feature extractors are fixed and a multi-class classifier is established and to fuse all mode features for high-precision multi-class identification via ensemble learning. As a variant of signal-processing-collaborated deep learning frameworks, the FIR-CCF method fully exploits the strengths of signal-processing methods in interpretability and feature extraction. Three experimental cases highlight the superiority and significant improvement of the FIR-CCF framework over other five state-of-the-art diagnosis methods and three ablation models. Specially, extensive visualization is implemented to place in-depth insight into how the FIR-CCF framework works. It can be also foreseen that the signal-processing-collaborated deep learning framework shows enormous potential in interpretable fault diagnosis for knowledge-informed artificial intelligence. Related source codes will be available at: https://github.com/BITS/FIR-CCF-main.

Evaluation of a Task-Specific Self-Supervised Learning Framework in Digital Pathology Relative to Transfer Learning Approaches and Existing Foundation Models

An integral stage in typical digital pathology workflows involves deriving specific features from tiles extracted from a tessellated whole-slide image. Notably, various computer vision neural network architectures, particularly the ImageNet pretrained, have been extensively used in this domain. This study critically analyzes multiple strategies for encoding tiles to understand the extent of transfer learning and identify the most effective approach. The study categorizes neural network performance into 3 weight initialization methods: random, ImageNet-based, and self-supervised learning. Additionally, we propose a framework based on task-specific self-supervised learning, which introduces a shallow feature extraction method, employing a spatial-channel attention block to glean distinctive features optimized for histopathology intricacies. Across 2 different downstream classification tasks (patch classification and weakly supervised whole-slide image classification) with diverse classification data sets, including colorectal cancer histology, Patch Camelyon, prostate cancer detection, The Cancer Genome Atlas, and CIFAR-10, our task-specific self-supervised encoding approach consistently outperforms other convolutional neural network–based encoders. The better performances highlight the potential of task-specific attention-based self-supervised training in tailoring feature extraction for histopathology, indicating a shift from using pretrained models originating outside the histopathology domain. Our study supports the idea that task-specific self-supervised learning allows domain-specific feature extraction, encouraging a more focused analysis.

Improved semantic segmentation method for weld penetration prediction of TIG welding with dual ellipsoid heat source

TIG welding is a manufacturing process that uses argon as a shielding gas and tungsten as an electrode to join metals at high temperatures. The weld penetration is the key to determining the quality of the weld. However, the lack of sensing technology makes weld penetration difficult to predict, which imposes a major challenge to process stability and weld quality. To address this challenge, a semantic segmentation-based approach for calibrating the weld penetration prediction model under dual ellipsoidal heat source is proposed to improve the prediction accuracy. To realize this, a semantic segmentation model Self-Attention-Structural-Reparameterization-BiSeNet (SASR-BiSeNet) is built for molten pool parameter extraction. The model introduces a self-attention mechanism as an enhancement to the convolution module. Meanwhile, the RepVGG-style structural reparameterization design decouples the training-time and inference-time architectures. Further, a mask feature extraction method is designed to obtain the calibration parameters of the molten pool and to correct the prediction model. The accuracy of the calibrated prediction model is verified by welding experiments using 304L steel materials on a robotic welding system. The results show that the maximum Mean Intersection over Union (mIoU) of the SASR-BiSeNet is 92.51 %, which is a 13.87 % improvement compared to the BiSeNet, and the maximum depth of molten pool error decreasing from 16.59 % to 9.84 %. This method can improve the precision of welding penetration control and plays an important role in promoting welding intelligence.

Optimized polycystic ovarian disease prognosis and classification using AI based computational approaches on multi-modality data

Polycystic Ovarian Disease or Polycystic Ovary Syndrome (PCOS) is becoming increasingly communal among women, owing to poor lifestyle choices. According to the research conducted by National Institutes of Health, it has been observe that PCOS, an endocrine condition common in women of childbearing age, has become a significant contributing factor to infertility. Ovarian abnormalities brought on by PCOS carry a high risk of miscarriage, infertility, cardiac problems, diabetes, uterine cancer, etc. Ovarian cysts, obesity, menstrual irregularities, elevated amounts of male hormones, acne vulgaris, hair loss, and hirsutism are some of the symptoms of PCOS. It is not easy to determine PCOS because of its different combinations of symptoms in different women and various criteria needed for diagnosis. Taking biochemical tests and ovary scanning is a time-consuming process and the financial expenses have become a hardship to the patients. Thus, early prognosis of PCOS is crucial to avoid infertility. The goal of the proposed work is to analyse PCOS symptoms based on clinical data for early diagnosis and to classify into PCOS affected or not. To achieve this objective, clinical features dataset and ultrasound imaging dataset from Kaggle is utilized. Initially 541 instances of 45 clinical features such as testosterone, hirsutism, family history, BMI, fast food, menstrual disorder, risk etc. are considered and correlation-based feature extraction method is applied to this dataset which results in 17 features. The extracted features are applied to various machine learning algorithms such as Logistic Regression, Naïve Bayes and Support Vector Machine. The performance of each method is evaluated based on accuracy, precision, recall, F1-score and the result shows that among three models, Support Vector Machine model achieved high accuracy of 94.44%. In addition to this, 3856 ultrasound images are analysed by CNN based deep learning algorithm and VGG16 transfer learning algorithm. The performance of these models is evaluated using training accuracy, loss and validation accuracy, loss and the result depicts that VGG16 outperforms than CNN model with validation accuracy of 98.29%.

Adaptive prototype few-shot image classification method based on feature pyramid

Few-shot learning aims to enable machines to recognize unseen novel classes using limited samples akin to human capabilities. Metric learning is a crucial approach to addressing this challenge, with its performance primarily dependent on the effectiveness of feature extraction and prototype computation. This article introduces an Adaptive Prototype few-shot image classification method based on Feature Pyramid (APFP). APFP employs a novel feature extraction method called FResNet, which builds upon the ResNet architecture and leverages a feature pyramid structure to retain finer details. In the 5-shot scenario, traditional methods for computing average prototypes exhibit limitations due to the typically diverse and uneven distribution of samples, where simple means may inadequately reflect such diversity. To address this issue, APFP proposes an Adaptive Prototype method (AP) that dynamically computes class prototypes of the support set based on the similarity between support set samples and query samples. Experimental results demonstrate that APFP achieves 67.98% and 85.32% accuracy in the 5-way 1-shot and 5-way 5-shot scenarios on the MiniImageNet dataset, respectively, and 84.02% and 94.44% accuracy on the CUB dataset. These results indicate that the proposed APFP method addresses the few-shot learning problem.

A comprehensive review of combating EDoS attacks in cloud services with deep learning and advanced network security technologies including DDoS protection and intrusion prevention systems

This comprehensive review examines the current state of research and practice in combating Economic Denial of Sustainability (EDoS) attacks in cloud services, with a focus on deep learning approaches and advanced network security technologies. The paper provides an in-depth analysis of EDoS attack characteristics, their impact on cloud economics, and the challenges faced in mitigation efforts. It explores the application of deep learning techniques in EDoS detection and prevention, highlighting recent advancements in neural network architectures and feature extraction methods. The review also covers the integration of advanced network security technologies, including next-generation firewalls, software-defined networking, and cloud-native security solutions, in the context of EDoS protection. Furthermore, it discusses the adaptation of Distributed Denial of Service (DDoS) mitigation strategies for EDoS attacks, emphasizing traffic analysis and anomaly detection techniques. The role of Intrusion Prevention Systems (IPS) in EDoS mitigation is examined, comparing signature-based and behavior-based approaches and exploring their integration with other security components. The paper concludes by identifying emerging threats, regulatory considerations, and open research problems in EDoS protection, providing valuable insights for researchers and practitioners in the field of cloud security. This review aims to serve as a comprehensive resource for understanding the current landscape of EDoS attacks and defense mechanisms, while also highlighting future directions for research and development in this critical area of cloud computing security.

A Review on Deepfake generation and Detection based on Deep learning: Approaches, and Future Challenges

In recent years, applications of deepfake, particularly to achieve political, economic, or social reputation aims, have been become widespread. These applications do not require high-level professional technical skills. Also, deep learning techniques like Generative Adversarial networks (GANs) have enhanced deepfake, making it more realistic. So, several researchers are looking for developing an effective method to detect a fake image or video. This paper provides a comprehensive overview of several proposed deepfake generation approaches and the approaches used to detect any manipulation. Based on feature extraction methods, this study provides an extensive review of face manipulation, especially focusing on facial swap, re-enactment, and attribute manipulation. Additionally, the study describes all existing deepfake methods and evaluates the presented detection models based on the most effective deep learning algorithms by comparing their respective evaluation metrics. Moreover, it presents the challenges and gapes in trying to enhance and develop deepfake detection techniques. It assists readers in understanding the generation and detection of deepfake mechanisms and presents the field limitations and future works.

Study on detection of pesticide residues in tobacco based on hyperspectral imaging technology.

Tobacco is a critical economic crop, yet its cultivation heavily relies on chemical pesticides, posing health risks to consumers, therefore, monitoring pesticide residues in tobacco is conducive to ensuring food safety. However, most current research on pesticide residue detection in tobacco relies on traditional chemical methods, which cannot meet the requirements for real-time and rapid detection. This study introduces an advanced method that combines hyperspectral imaging (HSI) technology with machine learning algorithms. Firstly, a hyperspectral imager was used to obtain spectral data of tobacco samples, and a variety of spectral pre-processing technologies such as mean centralization (MC), trend correction (TC), and wavelet transform (WT), as well as feature extraction methods such as competitive adaptive reweighted sampling (CARS) and least angle regression (LAR) were used to process the spectral data, and then, grid search algorithm (GSA) is used to optimize the support sector machine (SVM). The optimized MC-LAR-SVM model achieved a pesticide classification accuracy of 84.1%, which was 9.5% higher than the original data model. The accuracy of the WT-TC-CARS-GSA-SVM model in the fenvalerate concentration classification experiment was as high as 91.8 %, and it also had excellent performance in other metrics. Compared with the model based on the original data, the accuracy, precision, recall, and F1-score are improved by 8.3 %, 8.2 %, 7.5 %, and 0.08, respectively. The results show that combining spectral preprocessing and feature extraction algorithms with machine learning models can significantly enhance the performance of pesticide residue detection models and provide robust, efficient, and accurate solutions for food safety monitoring. This study provides a new technical means for the detection of pesticide residues in tobacco, which is of great significance for improving the efficiency and accuracy of food safety detection.

Robust Face Recognition System Based on Deep Facial Feature Extraction and Machine Learning

Facial Recognition is a common challenge in research for security applications. Wide-ranging applications of this technology include biometrics, security data, access to controlled locations, smart cards, and surveillance systems. A practical approach to identifying individuals depends on factors such as partial face occlusion, illumination, age, expression, makeup, and poses. These complex face recognition variables affect most face recognition systems. In this paper, we propose a deep transfer learning for facial features based on the pose and illumination variations of celebrities’ faces in a free environment for face recognition. The new approach that combines deep learning methods for feature extraction with machine learning classification methods. Also, studies the performance of the pre-trained model (VGGFace2 with ResNet-50 model weights) for feature extraction with a Multilayer Perceptron Classifier (MLP), Decision Tree, and Bootstrap Aggregation (Bagging) to perform classification. The convolution neural network has lately achieved excellent advancement in facial recognition. The outcome indicated that VGGFace2 with MLP obtained more precision and provided the best results (Accuracy, F-measure, Recall, and Precision). The results for all metric for the proposed model is 99%. The benchmark for this model is a dataset of 105 people’s faces.

Audio Signal Recognition in Complex Environments Using Sparse Representation

Recognizing audio signals in complex environments is crucial for acquiring adequate information. This paper integrates the sparse expression algorithm with Mel-frequency cepstral coefficient (MFCC) features. The combined approach was applied in convolutional neural network classifiers to recognize acoustic scenes in audio signals within complex environments. The algorithm was then simulated and tested using the TUT Sound Events 2016 and TUT Acoustic Scenes 2016 datasets. In the experiments, the efficacy of the developed sparse feature extraction method was validated. Then, the appropriate sparse dictionary size was determined. The algorithm was subsequently compared with two recognition algorithms based on sparse and MFCC features, respectively. It was found that the extraction approach proposed in this paper had a higher signal-to-noise ratio. The results revealed variations in the required sparse dictionary size for different datasets: 75 for TUT Sound Events 2016 and 150 for TUT Acoustic Scenes 2016. The MFCC-combined recognition algorithm demonstrated the fastest convergence during training among the three audio scene recognition algorithms. For both the TUT Sound Events 2016 and TUT Acoustic Scenes 2016 datasets, the MFCC-combined recognition algorithm achieved the highest classification accuracy, and the recognition accuracy for the former dataset was higher.

A feature extraction method for intelligent chatter detection in the milling process

A feature extraction method for intelligent chatter detection in the milling process

Information enhanced slow feature analysis integrated with prior fault data for sensitive monitoring of chemical processes

As an effective feature extraction method, slow feature analysis (SFA) has been successfully applied in the domain of chemical process monitoring. However, as an unsupervised method, the basic SFA ignores the key information carried by prior fault data so that many complicated faults cannot be sensitively detected. To address this issue, an enhanced SFA method, termed Information Enhanced SFA (IE-SFA), is proposed by integrating available fault discriminant information to achieve more sensitive detection of complicated chemical process faults. The proposed method builds a primary-auxiliary SFA modeling framework. On the one hand, the normal training data are analyzed to establish the primary SFA model. On the other, prior fault data are integrated to develop an auxiliary monitoring model for providing extra fault-sensitive information. In the auxiliary model, a sample-variable weighted Fisher discriminant analysis method is performed on normal data, fault data, and their respective slow feature components, thereby extracting fault-sensitive features for enhancing the fault detection capability. For full-view fault monitoring, the Bayesian fusion strategy is used to fuse the outcomes from both primary and auxiliary models. The applications on a benchmark Tennessee Eastman chemical process and a three-phase flow process demonstrate that the proposed IE-SFA method provides superior fault detection performance compared to the basic SFA method.

Optimizing Mail Sorting with Naive Bayes Classifier and Enhanced Feature Extraction Method

Optimizing Mail Sorting with Naive Bayes Classifier and Enhanced Feature Extraction Method

Optimizing Prostate Cancer Radiotherapy: Advanced Machine Learning in Virtual Patient-Specific Plan Verification-Review Study

Background: Virtual patient-specific plan verification in radiotherapy is critical to ensure precise dose delivery while minimizing healthy tissue exposure, particularly in prostate cancer treatments. Prior studies, however, have overlooked the physical implications of predictor features and lacked comprehensive decision support tools, leading to gaps in understanding and practical application. This research aims to bridge these gaps by providing a nuanced understanding of predictor features, exploring advanced automatic feature extraction methods, emphasizing model reliability, and proposing a comprehensive decision support tool. Our objective is to optimize radiotherapy protocols, ensuring safer and more effective treatments for patients undergoing prostate cancer therapy. Methods: This research employs a literature review approach. An extensive literature study served as the foundation. Results: Our study reveals that understanding the physical implications of predictor features significantly enhances prediction accuracy. Utilizing Convolutional Neural Networks (CNN) models for automatic feature extraction improves prediction performance, providing robust and transferable results. Conclusions: By emphasizing model reliability through the integration of treatment plan parameters, our approach ensures stable predictions across diverse patient cases. The proposed decision support tool offers clinicians detailed insights into predicted dose deliverability, facilitating informed decision-making for patient-specific treatment plans. Through these advancements, our research contributes to the optimization of radiotherapy protocols, ensuring safer and more effective treatments for patients undergoing prostate cancer therapy.

Fault Feature Extraction of Rolling Bearing Based on Maximum Second-Order Cyclostationarity Blind Deconvolution and CEEMD-Teager

The operation of rolling bearings is directly related to the reliability of the whole rotating machinery. If the rolling bearing failure cannot be diagnosed and repaired in time, it will probably lead to equipment shutdown. The feature extraction process is an important part of bearing fault diagnosis. Some existing feature extraction methods are sensitive to noise and interference, and cannot fully explore and characterize useful information in nonlinear and non-stationary signals in complex scenes, and sometimes even lose important information contained in the data, resulting in low diagnostic accuracy. Therefore, a new method combining Maximum Second-order Cyclostationarity Blind Deconvolution (CYCBD) and Complementary Ensemble Empirical Mode Decomposition (CEEMD) was proposed for feature extraction of rolling bearing faults. Firstly, the cyclic frequency is set according to the failure frequency, and the original signal is filtered by CYCBD. Then, the filtered signal is decomposed into multiple Intrinsic Mode Function (IMF) by CEEMD, and the effective IMF components are selected by kurtosis criterion for reconstruction. Finally, the Teager energy operator is used to enhance the transient impact of the reconstructed signals. The results of simulation and comparison experiments show that the proposed method can extract bearing characteristics in different fault states more effectively, and the accuracy of network model diagnosis is improved compared with the traditional methods.