Classification Ability Research Articles

Class Incremental Deep Learning: A Computational Scheme to Avoid Catastrophic Forgetting in Domain Generation Algorithm Multiclass Classification

Domain Generation Algorithms (DGAs) are algorithms present in most malware used by botnets and advanced persistent threats. These algorithms dynamically generate domain names to maintain and obfuscate communication between the infected device and the attacker’s command and control server. Since DGAs are used by many threats, it is extremely important to classify a given DGA according to the threat it is related to. In addition, as new threats emerge daily, classifier models tend to become obsolete over time. Deep neural networks tend to lose their classification ability when retrained with a dataset that is significantly different from the initial one, a phenomenon known as catastrophic forgetting. This work presents a computational scheme composed of a deep learning model based on CNN and natural language processing and an incremental learning technique for class increment through transfer learning to classify 60 DGA families and include a new family to the classifier model, training the model incrementally using some examples from known families, avoiding catastrophic forgetting and maintaining metric levels. The proposed methodology achieved an average precision of 86.75%, an average recall of 83.06%, and an average F1 score of 83.78% with the full dataset, and suffered minimal losses when applying the class increment.

Non-magnetic metal plate classification algorithm that can ignore the damage condition

Aiming at the problem of classification and recovery of non-magnetic metals, this paper proposes a method of classification of non-magnetic metals based on eddy current detection, which takes the slope of inductance real part as the conductivity characterization parameter to measure the conductivity of non-magnetic metals, to distinguish metals. The most significant advantage of this method is that even in the face of a metal plate with a damaged surface, its classification ability still exists, and the parameter of the slope of inductance real part will monotonically change with the surface area or depth of damage, which has guiding significance for damage classification or damage assessment.

Image Classification Based on Low-Level Feature Enhancement and Attention Mechanism

Deep learning-based image classification networks heavily rely on the extracted features. However, as the model becomes deeper, important features may be lost, resulting in decreased accuracy. To tackle this issue, this paper proposes an image classification method that enhances low-level features and incorporates an attention mechanism. The proposed method employs EfficientNet as the backbone network for feature extraction. Firstly, the Feature Enhancement Module quantifies and statistically processes low-level features from shallow layers, thereby enhancing the feature information. Secondly, the Convolutional Block Attention Module enhances the high-level features to improve the extraction of global features. Finally, the enhanced low-level features and global features are fused to supplement low-resolution global features with high-resolution details, further improving the model’s image classification ability. Experimental results illustrate that the proposed method achieves a Top-1 classification accuracy of 86.49% and a Top-5 classification accuracy of 96.90% on the ETH-Food101 dataset, 86.99% and 97.24% on the VireoFood-172 dataset, and 70.99% and 92.73% on the UEC-256 dataset. These results demonstrate that the proposed method outperforms existing methods in terms of classification performance.

Dynamic Feature Focusing Network for small object detection

Deep learning has driven research in object detection and achieved proud results. Despite its significant advancements in object detection, small object detection still struggles with low recognition rates and inaccurate positioning, primarily attributable to their miniature size. The location deviation of small objects induces severe feature misalignment, and the disequilibrium between classification and regression tasks hinders accurate recognition. To address these issues, we propose a Dynamic Feature Focusing Network (DFFN), which contains a duo of crucial modules: Visual Perception Enhancement Module (VPEM) and Task Association Module (TAM). Drawing upon the deformable convolution and attention mechanism, the VPEM concentrates on sparse key features and perceives the misalignment via positional offset. We aggregate multi-level features at identical spatial locations via layer average operation for learning a more discriminative representation. Incorporating class alignment and bounding box alignment parts, the TAM promotes classification ability, refines bounding box regression, and facilitates the joint learning of classification and localization. We conduct diverse experiments, and the proposed method considerably enhances the small object detection performance on four benchmark datasets of MS COCO, VisDrone, VOC, and TinyPerson. Our method has improved by 3.4 and 2.2 in mAP and APs, making solid improvements on COCO. Compared to other classic detection models, DFFN exhibits a high level of competitiveness in precision.

A deep learning model for anti-inflammatory peptides identification based on deep variational autoencoder and contrastive learning

As a class of biologically active molecules with significant immunomodulatory and anti-inflammatory effects, anti-inflammatory peptides have important application value in the medical and biotechnology fields due to their unique biological functions. Research on the identification of anti-inflammatory peptides provides important theoretical foundations and practical value for a deeper understanding of the biological mechanisms of inflammation and immune regulation, as well as for the development of new drugs and biotechnological applications. Therefore, it is necessary to develop more advanced computational models for identifying anti-inflammatory peptides. In this study, we propose a deep learning model named DAC-AIPs based on variational autoencoder and contrastive learning for accurate identification of anti-inflammatory peptides. In the sequence encoding part, the incorporation of multi-hot encoding helps capture richer sequence information. The autoencoder, composed of convolutional layers and linear layers, can learn latent features and reconstruct features, with variational inference enhancing the representation capability of latent features. Additionally, the introduction of contrastive learning aims to improve the model's classification ability. Through cross-validation and independent dataset testing experiments, DAC-AIPs achieves superior performance compared to existing state-of-the-art models. In cross-validation, the classification accuracy of DAC-AIPs reached around 88%, which is 7% higher than previous models. Furthermore, various ablation experiments and interpretability experiments validate the effectiveness of DAC-AIPs. Finally, a user-friendly online predictor is designed to enhance the practicality of the model, and the server is freely accessible at http://dac-aips.online.

Few-shot learning with representative global prototype

Few-shot learning is often challenged by low generalization performance due to the model is mostly learned with the base classes only. To mitigate the above issues, a few-shot learning method with representative global prototype is proposed in this paper. Specifically, to enhance generalization to novel class, we propose a strategy for jointly training base and novel classes. This process produces prototypes characterizing the class information called representative global prototypes. Additionally, to avoid the problem of data imbalance and prototype bias caused by newly added categories of sparse samples, a novel sample synthesis method is proposed for augmenting more representative samples of novel class. Finally, representative samples and non-representative samples with high uncertainty are selected to enhance the representational and discriminative abilities of the global prototype. Intensive experiments have been conducted on two popular benchmark datasets, and the experimental results show that this method significantly improves the classification ability of few-shot learning tasks and achieves state-of-the-art performance.

DCNNLFS: A Dilated Convolutional Neural Network With Late Fusion Strategy for Intelligent Classification of Gastric Histopathology Images.

Gastric cancer has a high incidence rate, significantly threatening patients' health. Gastric histopathology images can reliably diagnose related diseases. Still, the data volume of histopathology images is too large, making misdiagnosis or missed diagnosis easy. The classification model based on deep learning has made some progress on gastric histopathology images. However, traditional convolutional neural networks (CNNs) generally use pooling operations, which will reduce the spatial resolution of the image, resulting in poor prediction results. The image feature in previous CNN has a poor perception of details. Therefore, we design a dilated CNN with a late fusion strategy (DCNNLFS) for gastric histopathology image classification. The DCNNLFS model utilizes dilated convolutions, enabling it to expand the receptive field. The dilated convolutions can learn the different contextual information by adjusting the dilation rate. The DCNNLFS model uses a late fusion strategy to enhance the classification ability of DCNNLFS. We run related experiments on a gastric histopathology image dataset to verify the excellence of the DCNNLFS model, where the three metrics Precision, Accuracy, and F1-Score are 0.938, 0.935, and 0.959.

Dual-energy computed tomography iodine quantification combined with laboratory data for predicting microvascular invasion in hepatocellular carcinoma: a two-centre study.

Microvascular invasion (MVI) is a recognized biomarker associated with poorer prognosis in patients with hepatocellular carcinoma. Dual-energy computed tomography (DECT) is a highly sensitive technique that can determine the iodine concentration (IC) in tumour and provide an indirect evaluation of internal microcirculatory perfusion. This study aimed to assess whether the combination of DECT with laboratory data can improve preoperative MVI prediction. This retrospective study enrolled 119 patients who underwent DECT liver angiography at 2 medical centres preoperatively. To compare DECT parameters and laboratory findings between MVI-negative and MVI-positive groups, Mann-Whitney U test was used. Additionally, principal component analysis (PCA) was conducted to determine fundamental components. Mann-Whitney U test was applied to determine whether the principal component (PC) scores varied across MVI groups. Finally, a general linear classifier was used to assess the classification ability of each PC score. Significant differences were noted (P < .05) in alpha-fetoprotein (AFP) level, normalized arterial phase IC, and normalized portal phase IC between the MVI groups in the primary and validation datasets. The PC1-PC4 accounted for 67.9% of the variance in the primary dataset, with loadings of 24.1%, 16%, 15.4%, and 12.4%, respectively. In both primary and validation datasets, PC3 and PC4 were significantly different across MVI groups, with area under the curve values of 0.8410 and 0.8373, respectively. The recombination of DECT IC and laboratory features based on varying factor loadings can well predict MVI preoperatively. Utilizing PCA, the amalgamation of DECT IC and laboratory features, considering diverse factor loadings, showed substantial promise in accurately classifying MVI. There have been limited endeavours to establish such a combination, offering a novel paradigm for comprehending data in related research endeavours.

Optimization of headspace high-capacity tool coupled to two-dimensional gas chromatography–mass spectrometry for mapping the volatile organic compounds of raw pistachios. A proof-of-concept on the classification ability by geographic origin

An optimized procedure for extracting and analyzing raw pistachio volatiles was developed through headspace sampling with high-capacity tools and subsequent analysis using comprehensive two-dimensional gas chromatography coupled with mass spectrometry. The examination of 18 pistachio samples belonging to different geographic areas led to the identification of a set of 99 volatile organic compounds (VOCs). Molecules were putatively identified using linear retention index, mass spectra similarity, and two-dimensional plot location. The impact of preprocessing and processing techniques on the aligned data matrix from a set of samples of different geographical origins, after removing contaminants, was evaluated. The combination of scaling with log-transformation, normalization with z-score, and data reduction with random forest machine learning algorithm generated a panel of 16 discriminatory VOC molecules. As a proof of concept, raw pistachios' VOC profile was employed for the first time to tentatively classify them based on their geographical origin.

Plasma cell-free DNA as a monitoring tool for high-risk pregnancies associated with antiphospholipid syndrome

IntroductionDespite thromboprophylaxis, women with antiphospholipid syndrome (APS) face high-risk pregnancies due to proinflammatory and prothrombotic states. This highlights the need for new monitoring and prognostic tools. Recent insights into the pathophysiological role of neutrophil activation and extracellular trap (NET) formation in this syndrome led to the exploration of plasma cell-free DNA (cfDNA), a derivative of NETosis, as a promising biomarker. Materials and methodscfDNA was isolated and quantified from plasma samples of healthy pregnant women (control group, HC) and women with APS (APS group). We assessed the physiological variability of cfDNA across the three trimesters in HC. Levels of cfDNA were compared between APS and HC by gestational trimester. ROC curve analysis was performed to evaluate the efficacy of cfDNA levels for classifying APS patients. Furthermore, cfDNA levels in pregnant women with APS with obstetric complications were compared to those from uncomplicated pregnancies. ResultsAmong HC, cfDNA significantly increased in the third trimester compared to the first and second. Elevated cfDNA levels in APS compared to HC were observed in the first and second trimesters. First-trimester cfDNA levels demonstrated the highest classification ability to discriminate between APS and HC patients (AUC: 0.906). Among APS, those with complicated pregnancies (fetal growth restriction, preeclampsia, placenta accreta) exhibited significantly elevated cfDNA levels in the second trimester. ConclusionsElevated levels of cfDNA in pregnant women with APS, particularly among those with obstetric complications, supports further investigation into the potential of cfDNA as a valuable tool in the obstetric management of women with APS.

An epilepsy classification based on FFT and fully convolutional neural network nested LSTM.

Epilepsy, which is associated with neuronal damage and functional decline, typically presents patients with numerous challenges in their daily lives. An early diagnosis plays a crucial role in managing the condition and alleviating the patients' suffering. Electroencephalogram (EEG)-based approaches are commonly employed for diagnosing epilepsy due to their effectiveness and non-invasiveness. In this study, a classification method is proposed that use fast Fourier Transform (FFT) extraction in conjunction with convolutional neural networks (CNN) and long short-term memory (LSTM) models. Most methods use traditional frameworks to classify epilepsy, we propose a new approach to this problem by extracting features from the source data and then feeding them into a network for training and recognition. It preprocesses the source data into training and validation data and then uses CNN and LSTM to classify the style of the data. Upon analyzing a public test dataset, the top-performing features in the fully CNN nested LSTM model for epilepsy classification are FFT features among three types of features. Notably, all conducted experiments yielded high accuracy rates, with values exceeding 96% for accuracy, 93% for sensitivity, and 96% for specificity. These results are further benchmarked against current methodologies, showcasing consistent and robust performance across all trials. Our approach consistently achieves an accuracy rate surpassing 97.00%, with values ranging from 97.95 to 99.83% in individual experiments. Particularly noteworthy is the superior accuracy of our method in the AB versus (vs.) CDE comparison, registering at 99.06%. Our method exhibits precise classification abilities distinguishing between epileptic and non-epileptic individuals, irrespective of whether the participant's eyes are closed or open. Furthermore, our technique shows remarkable performance in effectively categorizing epilepsy type, distinguishing between epileptic ictal and interictal states versus non-epileptic conditions. An inherent advantage of our automated classification approach is its capability to disregard EEG data acquired during states of eye closure or eye-opening. Such innovation holds promise for real-world applications, potentially aiding medical professionals in diagnosing epilepsy more efficiently.

A random feature mapping method based on the AdaBoost algorithm and results fusion for enhancing classification performance

The feature mapping method can improve data separability, enhance data representation ability, and reduce data processing complexity. However, on the one hand, the existing feature mapping methods have difficulty processing datasets of different dimensions and distributions adaptively, limiting the scope of application; on the other hand, a single feature mapping method has the problem of instability and poor generalization ability, weakening the classification ability of subsequent classifiers. This paper proposes a random feature mapping method based on the AdaBoost algorithm and results fusion to enhance classification performance. The method adopts horizontal expansion, fine-tuning weights through sparse autoencoders, and uses input-mapped features as feature nodes to generate multiple feature subsets for increasing stability. After training weak classifiers on each multiple feature subset, the weights of classifiers are adjusted adaptively by the Adaboost ensemble algorithm. Finally, the method fuses weak classifiers twice to enhance classification performance, which abandons the traditional voting method and uses the weighted probability selection method. Experiments on twenty classic datasets show that the proposed method can effectively mine essential features and enhance classification accuracy compared with original datasets. For instance, the Balance dataset has an average classification accuracy of more than 20% higher than the original dataset on the KNN classifier. The proposed method outperforms alternative feature mapping methods in terms of performance and efficiency on different classifiers in most cases.

Effects of Wall and Freespace Damping Levels on Virtual Wall Stiffness Classification.

Virtual damping is often employed to improve stability in virtual environments, but it has previously been found to bias perception of stiffness, with its effects differing when it is introduced locally within a wall/object or globally in both the wall and in freespace. Since many potential applications of haptic rendering involve not only comparisons between two environments, but also the ability to recognize rendered environments as belonging to different categories, it is important to understand the perceptual impacts of freespace and wall damping on stiffness classification ability. This study explores the effects of varying levels of freespace and wall damping on users' ability to classify virtual walls by their stiffness. Results indicate that freespace damping improves wall classification if the walls are damped, but will impair classification of undamped walls. These findings suggest that, in situations where users are expected to recognize and classify various stiffnesses, freespace damping can be a factor in narrowing or widening gaps in extended rate-hardness between softer and stiffer walls.

Single sample-oriented attribute reduction for rule learning with formal concept analysis

As an effective tool for data mining, formal concept analysis can yield interpretable decision rules using attribute reduction. Currently, existing reduction methods within the framework of formal concept analysis result in all samples sharing the same conditional attributes, thereby overlooking the distinctions between rules extracted from different samples. To address this problem, we present a novel attribute reduction method for single samples. For this purpose, a single sample-oriented reduction framework is established by incorporating a discernibility attribute vector and an algorithm for calculating reducts based on the matrix representation of the discernibility attribute vector. Furthermore, the process of obtaining a new reduct from the original reduct is discussed by considering incremental learning. Given that attributes included in reducts are crucial conditions for decision-making, intent-minimal granular rules are generated through attribute reduction, and a reduction-based method for measuring the completeness of the rule set is discussed. Finally, to accomplish the classification task, a minimal rule-based classification model (MRCM) is proposed. The experimental results show that the MRCM has the same classification ability as the unreduced rules, and the rules with higher completeness tend to exhibit better classification performance.

A combination of left ventricular outflow tract velocity time integral and lung ultrasound to predict mortality in ST elevation myocardial infarction.

Development of ventricular failure and pulmonary edema is associated with a worse prognosis in ST-elevation myocardial infarction (STEMI). We aimed to evaluate the prognostic ability of a novel classification combining lung ultrasound (LUS) and left ventricular outflow tract (LVOT) velocity time integral (VTI) in patients with STEMI. LUS and LVOT-VTI were performed within 24h of admission in STEMI patients. A LUS combined with LVOT-VTI (LUV) classification was developed based on LUS with < or ≥ 3 positive zone scans, combined with LVOT-VTI > or ≤ 14. Patients were classified as A (< 3zones/ > 14cm VTI), B (≥ 3zones/ > 14cm VTI), C (< 3zones/ ≤ 14cm VTI) and D (≥ 3zones/ ≤ 14cm VTI). Primary outcome was occurrence of in-hospital mortality. Development of cardiogenic shock (CS) within 24h was also assessed. A total of 308 patients were included. Overall in-hospital mortality was 8.8%, while mortality for LUV A, B, C, and D was 0%, 3%, 12%, and 45%, respectively. The area under the curve (AUC) for predicting in-hospital mortality was 0.915. Moreover, after exclusion of patients admitted in Killip IV, at each increasing degree of LUV, a higher proportion of patients developed CS within 24h: LUV A = 0.0%, LUV B 5%, LUV C = 12.5% and LUV D = 30.8% (p < 0.0001). The AUC for predicting CS was 0.908 (p < 0.001). In a cohort of STEMI patients, LUV provided to be an excellent method for prediction of in-hospital mortality and development of CS. LUV classification is a fast, non-invasive and very user-friendly ultrasonographic evaluation method to stratify the risk of mortality and CS.

FutureCite: Predicting Research Articles’ Impact Using Machine Learning and Text and Graph Mining Techniques

The growth in academic and scientific publications has increased very rapidly. Researchers must choose a representative and significant literature for their research, which has become challenging worldwide. Usually, the paper citation number indicates this paper’s potential influence and importance. However, this standard metric of citation numbers is not suitable to assess the popularity and significance of recently published papers. To address this challenge, this study presents an effective prediction method called FutureCite to predict the future citation level of research articles. FutureCite integrates machine learning with text and graph mining techniques, leveraging their abilities in classification, datasets in-depth analysis, and feature extraction. FutureCite aims to predict future citation levels of research articles applying a multilabel classification approach. FutureCite can extract significant semantic features and capture the interconnection relationships found in scientific articles during feature extraction using textual content, citation networks, and metadata as feature resources. This study’s objective is to contribute to the advancement of effective approaches impacting the citation counts in scientific publications by enhancing the precision of future citations. We conducted several experiments using a comprehensive publication dataset to evaluate our method and determine the impact of using a variety of machine learning algorithms. FutureCite demonstrated its robustness and efficiency and showed promising results based on different evaluation metrics. Using the FutureCite model has significant implications for improving the researchers’ ability to determine targeted literature for their research and better understand the potential impact of research publications.

A brain tumor diagnosis approach based on deep separable convolutional networks and attention mechanisms

Brain tumors rank among the most lethal types of tumors. Magnetic Resonance Imaging (MRI) technology can clearly display the position, size, and borders of the tumors. Hence, MRI is frequently used in clinical diagnostics to detect brain tumors. Deep learning and related methods have been widely used in computer vision research recently for diagnosis and classification of MRI images of brain tumors. One trend is to achieve better performance by increasing model complexity. However, at the same time, the trainable parameters of the model increases accordingly. Too many parameters will lead to more difficult model training and optimization, and are prone to overfitting. To address this problem, this study constructs a model that incorporates a depthwise separable convolution technique and an attention mechanism to balance model performance and complexity. The model achieves 97.41% accuracy in the brain tumor classification task, which exceeds the 96.72% accuracy of the pre-trained model MobileNetV2, and shows good image classification ability. Future work could test the model's classification results on noisy images and investigate how to optimize the model's generalization ability.

A brain tumor diagnosis approach based on deep separable convolutional networks and attention mechanisms

Brain tumors rank among the most lethal types of tumors. Magnetic Resonance Imaging (MRI) technology can clearly display the position, size, and borders of the tumors. Hence, MRI is frequently used in clinical diagnostics to detect brain tumors. Deep learning and related methods have been widely used in computer vision research recently for diagnosis and classification of MRI images of brain tumors. One trend is to achieve better performance by increasing model complexity. However, at the same time, the trainable parameters of the model increases accordingly. Too many parameters will lead to more difficult model training and optimization, and are prone to overfitting. To address this problem, this study constructs a model that incorporates a depthwise separable convolution technique and an attention mechanism to balance model performance and complexity. The model achieves 97.41% accuracy in the brain tumor classification task, which exceeds the 96.72% accuracy of the pre-trained model MobileNetV2, and shows good image classification ability. Future work could test the model's classification results on noisy images and investigate how to optimize the model's generalization ability.

Multi-task disagreement-reducing multimodal sentiment fusion network

Existing multimodal sentiment analysis models can effectively capture sentimental commonalities between different modalities and possess high sentimental acquisition capability. However, there are still shortcomings in the model's analysis and recognition abilities when dealing with samples that exhibit sentimental polarity disagreement between different modalities. Additionally, the dominance of the text modality in multimodal models, particularly those pre-trained with BERT, can hinder the learning of other modalities due to its richer semantic information. This issue becomes particularly pronounced in cases where there is a conflict between multimodal and textual sentimental polarities, often leading to suboptimal analytical results. Besides, the classification ability of each modality is also suppressed by single-task learning. In this paper, We propose a Multi-Task disagreement-Reducing Multimodal Sentiment Fusion Network (MtDr-MSF), designed to enhance the semantic information of non-text modalities and reduce the dominant impact of the textual modality on the model, and to improve the learning capabilities of unimodal networks. We conducted experiments on multimodal sentiment analysis datasets, CMU-MOSI, CMU-MOSEI, and CH-SIMS. The results show that our method outperforms the current SOTA method.

Rolling bearing fault diagnosis method based on PE-DCM and ViT

Abstract Considering the issue of capturing the local and global contextual information and enhancing the parallel capability of bearing fault diagnosis in variable load and noise environments, a fault diagnosis method of rolling bearing based on PE-DCM and ViT is proposed. Firstly, the one-dimensional vibration signal is converted into a two-dimensional time-frequency diagram by continuous wavelet transform in the data processing module, and the model can understand the characteristics of the vibration signal more comprehensively. Secondly, a pyramid exponential expansion convolution module is established to extract the local features of fault information. Then, the global features of the fault information are learnt through the ViT (Vision Transformer) network, and the adaptive multi-attention is used to dynamically adjust the attention weights according to the features of the input data so as to inhibit noise or unimportant information. Finally, the experimental verification is carried out by using Case Western Reserve University and self-made MFS-bearing data set. The experimental results show that the method can better reflect the powerful image classification ability of the ViT network and has better noise resistance and generalization compared with other fault diagnosis methods.&amp;#xD;