SVDF: enhancing structural variation detect from long-read sequencing via automatic filtering strategies.

Structural variation (SV) is an important form of genomic variation that influences gene function and expression by altering the structure of the genome. Although long-read data have been proven to better characterize SVs, SVs detected from noisy long-read data still include a considerable portion of false-positive calls. To accurately detect SVs in long-read data, we present SVDF, a method that employs a learning-based noise filtering strategy and an SV signature-adaptive clustering algorithm, for effectively reducing the likelihood of false-positive events. Benchmarking results from multiple orthogonal experiments demonstrate that, across different sequencing platforms and depths, SVDF achieves higher calling accuracy for each sample compared to several existing general SV calling tools. We believe that, with its meticulous and sensitive SV detection capability, SVDF can bring new opportunities and advancements to cutting-edge genomic research.

Urine output is an early and strong predictor of acute kidney injury and associated mortality: a systematic literature review of 50 clinical studies.

BACKGROUND: Although the present diagnosis of acute kidney injury (AKI) involves measurement of acute increases in serum creatinine (SC) and reduced urine output (UO), measurement of UO is underutilized for diagnosis of AKI in clinical practice. The purpose of this investigation was to conduct a systematic literature review of published studies that evaluate both UO and SC in the detection of AKI to better understand incidence, healthcare resource use, and mortality in relation to these diagnostic measures and how these outcomes may vary by population subtype. METHODS: The systematic literature review was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist. Data were extracted from comparative studies focused on the diagnostic accuracy of UO and SC, relevant clinical outcomes, and resource usage. Quality and validity were assessed using the National Institute for Health and Care Excellence (NICE) single technology appraisal quality checklist for randomized controlled trials and the Newcastle-Ottawa Quality Assessment Scale for observational studies. RESULTS: A total of 1729 publications were screened, with 50 studies eligible for inclusion. A majority of studies (76%) used the Kidney Disease: Improving Global Outcomes (KDIGO) criteria to classify AKI and focused on the comparison of UO alone versus SC alone, while few studies analyzed a diagnosis of AKI based on the presence of both UO and SC, or the presence of at least one of UO or SC indicators. Of the included studies, 33% analyzed patients treated for cardiovascular diseases and 30% analyzed patients treated in a general intensive care unit. The use of UO criteria was more often associated with increased incidence of AKI (36%), than was the application of SC criteria (21%), which was consistent across the subgroup analyses performed. Furthermore, the use of UO criteria was associated with an earlier diagnosis of AKI (2.4-46.0 h). Both diagnostic modalities accurately predicted risk of AKI-related mortality. CONCLUSIONS: Evidence suggests that the inclusion of UO criteria provides substantial diagnostic and prognostic value to the detection of AKI.

Prospective investigation of amino acid transport and PSMA-targeted positron emission tomography for metastatic lobular breast carcinoma.

PURPOSE: To explore the feasibility of imaging amino-acid transport and PSMA molecular pathways in the detection of metastatic breast invasive lobular carcinoma (ILC) and if there is superior detection compared to standard-of-care imaging [computed tomography (CT)/bone scan, or 18F-FDG positron-emission-tomography (PET)-CT]. METHODS: 20 women with de-novo or suspected metastatic ILC underwent two PET-CT scans with 18F-fluciclovine and 68Ga-PSMA-11 on separate days. Uptake per patient and in 3 regions per patient - ipsilateral axillary lymph node (LN), extra-axillary LN (ipsilateral supraclavicular or internal mammary), or distant sites of disease - was compared to standard-of-care imaging (CT/bone scan in 13 patients and 18F-FDG PET-CT in 7 patients). Results were correlated to a composite standard of truth. Confirmed detection rate (cDR) was compared using McNemar's test. Mean SUVmax of 18F-fluciclovine and 68Ga-PSMA-11 in the most avid lesion for each true positive metastatic region and intact primary lesion were compared by t-test. RESULTS: The cDR for standard-of-care imaging was 5/20 patients in 5/60 regions. 68Ga-PSMA-11 PET-CT detected metastasis in 7/20 patients in 7/60 regions. 18F-fluciclovine PET-CT detected metastasis in 9/20 patients in 12/60 regions. The cDR for 18F-fluciclovine PET-CT was significantly higher versus standard-of-care imaging on the patient and combined region levels, while there were no significant differences between 68Ga-PSMA-11 and standard-of care imaging. 18F-fluciclovine cDR was also significantly higher than 68Ga-PSMA-11 on the combined region level. Mean SUVmax for true positive metastatic and primary lesions with 18F-fluciclovine (n = 18) was significantly greater than for 68Ga-PSMA-11 (n = 11) [5.5 ± 1.8 versus 3.5 ± 2.7 respectively, p = 0.021]. CONCLUSION: In this exploratory trial, 18F-fluciclovine PET-CT has a significantly higher cDR for ILC metastases compared to standard-of-care imaging and to 68Ga-PSMA-11. Mean SUVmax for true positive malignancy was significantly higher with 18F-fluciclovine than for 68Ga-PSMA-11. Exploratory data from this trial suggests that molecular imaging of amino acid metabolism in patients with ILC deserves further study. CLINICAL TRIAL REGISTRATION: Early phase (I-II) clinical trial (NCT04750473) funded by the National Institutes of Health (R21CA256280).

Basic research for identification and classification of organophosphorus pesticides in water based on ultraviolet-visible spectroscopy information.

In this study, the goal was to develop a method for detecting and classifying organophosphorus pesticides (OPPs) in bodies of water. Sixty-five samples with different concentrations were prepared for each of the organophosphorus pesticides, namely chlorpyrifos, acephate, parathion-methyl, trichlorphon, dichlorvos, profenofos, malathion, dimethoate, fenthion, and phoxim, respectively. Firstly, the spectral data of all the samples was obtained using a UV-visible spectrometer. Secondly, five preprocessing methods, six manifold learning methods, and five machine learning algorithms were utilized to build detection models for identifying OPPs in water bodies. The findings indicate that the accuracy of machine learning models trained on data preprocessed using convolutional smoothing + first-order derivatives (SG + FD) outperforms that of models trained on data preprocessed using other methods. The backpropagation neural network (BPNN) model exhibited the highest accuracy rate at 99.95%, followed by the support vector machine (SVM) and convolutional neural network (CNN) models, both at 99.92%. The extreme learning machine (ELM) and K-nearest neighbors (KNN) models demonstrated accuracy rates of 99.84% and 99.81%, respectively. Following the application of a manifold learning algorithm to the full-wavelength data set for the purpose of dimensionality reduction, the data was then visualized in the first three dimensions. The results demonstrate that the t-distributed domain embedding (t-SNE) algorithm is superior, exhibiting dense clustering of similar clusters and clear classification of dissimilar ones. SG + FD-t-SNE-SVM ranks highest among the feature extraction models in terms of performance. The feature extraction dimension was set to 4, and the average classification accuracy was 99.98%, which slightly improved the prediction performance over the full-wavelength model. As shown in this study, the ultraviolet-visible (UV-visible) spectroscopy system combined with the t-SNE and SVM algorithms can effectively identify and classify OPPs in waterbodies.

Traditional to technological advancements in Ganoderma detection methods in oil palm.

Ganoderma sp., the fungal agent causing basal stem rot (BSR), poses a severe threat to global oil palm production. Alarming increases in BSR occurrences within oil palm growing zones are attributed to varying effectiveness in its current management strategies. Asymptomatic progression of the disease and the continuous monoculture of oil palm pose challenges for prompt and effective management. Therefore, the development of precise, early, and timely detection techniques is crucial for successful BSR management. Conventional methods such as visual assessments, culture-based assays, and biochemical and physiological approaches prove time-consuming and lack specificity. Serological-based diagnostic methods, unsuitable for fungal diagnostics due to low sensitivity, assay affinity, cross-contamination which further underscores the need for improved techniques. Molecular PCR-based assays, utilizing universal, genus-specific, and species-specific primers, along with functional primers, can overcome the limitations of conventional and serological methods in fungal diagnostics. Recent advancements, including real-time PCR, biosensors, and isothermal amplification methods, facilitate accurate, specific, and sensitive Ganoderma detection. Comparative whole genomic analysis enables high-resolution discrimination of Ganoderma at the strain level. Additionally, omics tools such as transcriptomics, proteomics, and metabolomics can identify potential biomarkers for early detection of Ganoderma infection. Innovative on-field diagnostic techniques, including remote methods like volatile organic compounds profiling, tomography, hyperspectral and multispectral imaging, terrestrial laser scanning, and Red-Green-Blue cameras, contribute to a comprehensive diagnostic approach. Ultimately, the development of point-of-care, early, and cost-effective diagnostic techniques accessible to farmers is vital for the timely management of BSR in oil palm plantations.

Modified glove removal technique to prevent hand contamination in routine phlebotomy.

The World Health Organization and the Centers for Disease Control and Prevention (CDC) have established guidelines recommending the performance of hand hygiene routines for healthcare workers following glove removal. However, the completion of frequent hygiene routines can cause allergic and adverse skin reactions. This double-blind, randomized study aimed to address this concern by developing and evaluating a modified glove removal technique that minimizes contamination risk during routine phlebotomy procedures. Furthermore, this study used fluorescent detection to compare the frequency of contamination associated with the CDC-recommended technique and the modified technique using fluorescent detection. One hundred healthcare personnel were enrolled and divided into two groups: one group followed the CDC technique, while the other group implemented the modified technique. Participants received instructional videos and practiced under supervision. They subsequently performed blood collection using a simulation arm covered with fluorescent cream as a contamination marker. After removing gloves, hand contamination was assessed under a black light. The median time required for glove removal in the modified group was four seconds longer than that in the group that followed the CDC technique (p < 0.001). Contamination was observed in 2% (1/50) of subjects using the CDC-recommended technique, while no contamination was detected with the modified technique (p ≥ 0.05). Both the group that followed the CDC technique and the group that used modified glove removal techniques demonstrated the potential to prevent contamination during phlebotomy, thereby reducing the need for hand hygiene and the occurrence of contamination and adverse skin reactions. These findings prompt further exploration into whether proper glove removal can reduce the frequency of completing a hand hygiene routine after each glove removal, specifically within the context of phlebotomy. However, it is essential to note that hand hygiene following glove removal is still recommended to prevent contamination. Further research is warranted to validate these findings.

Establishment of a multi-line immunochromatography based on magnetic nanoparticles for simultaneous screening of multiple biomarkers.

Biomarkers screening is a benefit approach for early diagnosis of major diseases. In this study, magnetic nanoparticles (MNPs) have been utilized as labels to establish a multi-line immunochromatography (MNP-MLIC) for simultaneous detection of carcinoembryonic antigen (CEA), carbohydrate antigen 199 (CA 19-9), and alpha-fetoprotein (AFP) in a single serum sample. Under the optimal parameters, the three biomarkers can be rapidly and simultaneously qualitative screening within 15 min by naked eye. As for quantitative detection, the MNP-MLIC test strips were precisely positioned and captured by a smartphone, and signals on the test and control lines were extracted by ImageJ software. The signal ratio of test and control lines has been calculated and used to plot quantitative standard curves with the logarithmic concentration, of which the correlation coefficients are more than 0.99, and the limit of detection for CEA, CA 19-9, and AFP were 0.60 ng/mL, 1.21 U/mL, and 0.93 ng/mL, respectively. The recoveries of blank serum were 75.0 ~ 112.5% with the relative standard deviation ranging from 2.5 to 15.3%, and the specificity investigation demonstrated that the MNP-MLIC is highly specific to the three biomarkers. In conclusion, the developed MNP-MLIC offers a rapid, simple, accurate, and highly specific method for simultaneously detecting multiple biomarkers in serum samples, which provides an efficient and accurate approach for the early diagnosis of diseases.

GEB-YOLO: a novel algorithm for enhanced and efficient detection of foreign objects in power transmission lines.

Detecting foreign objects in power transmission lines is essential for mitigating safety risks and maintaining line stability. Practical detection, however, presents challenges including varied target sizes, intricate backgrounds, and large model weights. To address these issues, this study introduces an innovative GEB-YOLO model, which balances detection performance and quantification. Firstly, the algorithm features a lightweight architecture, achieved by merging the GhostConv network with the advanced YOLOv8 model. This integration considerably lowers computational demands and parameters through streamlined linear operations. Secondly, this paper proposes a novel EC2f mechanism, a groundbreaking feature that bolsters the model's information extraction capabilities. It enhances the relationship between weights and channels via one-dimensional convolution. Lastly, the BiFPN mechanism is employed to improve the model's processing efficiency for targets of different sizes, utilizing bidirectional connections and swift feature fusion for normalization. Experimental results indicate the model's superiority over existing models in precision and mAP, showing improvements of 3.7 and 6.8%, respectively. Crucially, the model's parameters and FLOPs have been reduced by 10.0 and 7.4%, leading to a model that is both lighter and more efficient. These advancements offer invaluable insights for applying laser technology in detecting foreign objects, contributing significantly to both theory and practice.

Night target detection algorithm based on improved YOLOv7.

Aiming at the problems of error detection and missing detection in night target detection, this paper proposes a night target detection algorithm based on YOLOv7(You Only Look Once v7). The algorithm proposed in this paper preprocesses images by means of square equalization and Gamma transform. The GSConv(Group Separable Convolution) module is introduced to reduce the number of parameters and the amount of calculation to improve the detection effect. ShuffleNetv2_×1.5 is introduced as the feature extraction Network to reduce the number of Network parameters while maintaining high tracking accuracy. The hard-swish activation function is adopted to greatly reduce the delay cost. At last, Scylla Intersection over Union function is used instead of Efficient Intersection over Union function to optimize the loss function and improve the robustness. Experimental results demonstrate that the average detection accuracy of the proposed improved YOLOv7 model is 88.1%. It can effectively improve the detection accuracy and accuracy of night target detection.

Evaluation and comparison of one-step real-time PCR and one-step RT-LAMP methods for detection of SARS-CoV-2.

BACKGROUND: There is an increasing disease trend for SARS-COV-2, so need a quick and affordable diagnostic method. It should be highly accurate and save costs compared to other methods. The purpose of this research is to achieve these goals. METHODS: This study analyzed 342 samples using TaqMan One-Step RT-qPCR and fast One-Step RT-LAMP (Reverse Transcriptase Loop-Mediated Isothermal Amplification). The One-Step LAMP assay was conducted to assess the sensitivity and specificity. RESULTS: The research reported positive samples using two different methods. In the RT-LAMP method, saliva had 92 positive samples (26.9%) and 250 negative samples (73.09%) and nasopharynx had 94 positive samples (27.4%) and 248 negative samples (72.51%). In the RT-qPCR method, saliva had 86 positive samples (25.1%) and 256 negative samples (74.8%) and nasopharynx had 93 positive samples (27.1%) and 249 negative samples (72.8%). The agreement between the two tests in saliva and nasopharynx samples was 93% and 94% respectively, based on Cohen's kappa coefficient (κ) (P < 0.001). The rate of sensitivity in this technique was reported at a dilution of 1 × 101 and 100% specificity. CONCLUSIONS: Based on the results of the study the One-Step LAMP assay has multiple advantages. These include simplicity, cost-effectiveness, high sensitivity, and specificity. The One-Step LAMP assay shows promise as a diagnostic tool. It can help manage disease outbreaks, ensure prompt treatment, and safeguard public health by providing rapid, easy-to-use testing.

A new diagnostic technique for identifying Angiostrongylus spp. larvae in intermediate snail species by examining the buccal cavity.

BACKGROUND: Angiostrongyliasis is a zoonotic parasitic disease caused by the rat lungworm Angiostrongylus cantonensis. The intermediate hosts of A. cantonensis are gastropods, and snail species such as Pomacea canaliculata play a key role in the transmission of human angiostrongyliasis. Detecting A. cantonensis infection in snails is an important component of epidemiological surveillance and the control of angiostrongyliasis. METHODS: In this study, a new method for diagnosing A. cantonensis infection in gastropods was developed by recovering larvae from the buccal cavity of three snail species. The entire buccal cavity of a snail was extracted, and the tissue was pressed between two microscope slides to observe whether A. cantonensis larvae were present. Our new method was compared with traditional pathogenic detection methods of lung microscopy, tissue homogenization, and artificial digestion. We artificially infected 160 P. canaliculata, 160 Cipangopaludina chinensis, and 160 Bellamya aeruginosa snails with A. cantonensis. Then, the four different detection methods were used to diagnose infection in each snail species at 7, 14, 21, and 28 days post exposure. RESULTS: We found no significant difference in the percentages of infected P. canaliculata snails using the four methods to detect A. cantonensis larvae. The radula pressing method had a mean detection rate of 80%, while the lung microscopy (81.3%), tissue homogenization (83.8%), and artificial digestion (85%) methods had slightly greater detection rates. Similarly, the percentages of infected C. chinensis snails that were detected using the radula pressing (80%), tissue homogenization (82.1%), and artificial digestion (83.8%) methods were not significantly different. Finally, the percentages of infected B. aeruginosa snails that were detected using the radula pressing (81.3%), tissue homogenization (81.9%), and artificial digestion (81.4%) methods were not significantly different. These results showed that the radula pressing method had a similar detection rate to traditional lung microscopy, tissue homogenization, or artificial digestion methods. CONCLUSIONS: This study demonstrates a new method for the qualitative screening of gastropods that act as intermediate hosts of A. cantonensis (and other Angiostrongylus species), provides technical support for the control of human angiostrongyliasis, and furthers research on A. cantonensis.

Establishing harmonised screening limits and detection times in greyhound racing - A considered approach.

An outline of the approach taken by international greyhound regulators to establish internationally harmonised screening limits and detection times in greyhound racing, which included a program of administration studies and an extensive and recognised risk assessment process, to ensure delivery of an effective anti-doping and medication control program.

First Report of Seed Decay Caused by Diaporthe ueckeri on soybean in Brazil.

A survey of Diaporthe/Phomopsis Complex (DPC) species was carried out on 479 asymptomatic soybean (Glycine max (L.) Merrill) seed samples collected from commercial soybean fields in the states of Santa Catarina (20 counties) and Rio Grande do Sul (41 counties), in the 2020/21 (n=186), 2021/22 (n=138) and 2022/23 (n=155) seasons from 120 cultivars. The seeds were provided by seed producers who collected according to the sampling standard of the Ministry of Agriculture, Livestock and Food Supply. From each sample received, 200 symptomless seeds were randomly sorted out. The seeds were surface disinfected by immersion in a sodium hypochlorite solution (1%) for two minutes and placed on Potato Dextrose Agar (PDA). The plates were incubated for 7 days at 23°C with a photoperiod of 12-h. The average prevalence of 73.7% of DPC-infected seeds. Colonies were isolated by transferring mycelial tips to PDA and incubating for 14 days at 25ºC in a 12-h photoperiod. One colony (isolate MEMR0500) had morphological characteristics similar to those reported in Lopez-Cardona (2021). This isolate had a floccose, dense colony ranging from grayish beige to brown with greenish regions and black globose pycnidia (3 to 4 pycnidia/cm²). Alpha-conidia, 5.1 to 7.0 µm x 1.5 to 2.8 µm, were observed after 30 days and were hyaline, aseptate and fusiform (Figure S1). No beta-conidia were observed. Soybean plants of cultivars BMX Cromo IPRO, BMX Zeus IPRO, BRS 5804 RR, FPS 1867 IPRO and NEO 750 IPRO were tested for pathogenicity using the toothpick inoculation method (Siviero and Menten 1995). Non-colonized toothpicks served as a negative control. Plants were incubated for four days at 25°C and 90% relative humidity. Elongated 1.0 to 2.5 cm x 0.5 to 0.9 cm lesions gray-brown/reddish-brown with a depressed center were observed in all inoculated cultivars. The fungus was reisolated and the characteristics of the colonies were identical to those previously isolated. For molecular characterization, DNA was extracted from the mycelia using the CTAB method (Doyle and Doyle 1990). End-point PCR was performed using GoTaq® Flexi DNA Polymerase (Promega, USA) and primer pairs, ITS-4F/ITS-5, T2/Bt2b and EF1-728F/EF1-986R to amplify the internal transcribed spacer (ITS) (Costamilan et al. 2008), ß-tubulin (TUB2) (Glass and Donaldson 1995), and translation elongation factor 1-α (TEF1) (Carbone and Kohn 1999) genes, respectively. The amplified fragments were sequenced and submitted to blast search (https://blast.ncbi.nlm.nih.gov/Blast.cgi) with the sequences available in GenBank. The fragment from ITS (accession number OR912979) showed 99.8% (549/582 bp) identity with Diaporthe ueckeri Udayanga & Castl. [as 'ueckerae'] [syn. D. miriciae R.G. Shivas, S.M. Thomps. & Y.P. Tan] isolate FAU656 (Ac. N. KJ590726). The sequence of TEF (Ac. N. PP372869) showed 99.7% (339/355 bp) identity with D. ueckeri FAU656 (Ac. N. KJ590747), and of TUB (Ac. N. PP372870) showed 98.9% (436/536 bp) identity with D. ueckeri FAU656 (Ac. N. KJ610881). A phylogenetic tree with amplified sequences of each gene and the corresponding representative sequences from the DPC was constructed in MEGA X (Kumar et al. 2018). The MEMR0500 isolate was clustered only with the D. ueckeri clade, confirming the identity of the fungus (Figure S2). In Brazil, this is the first report of the association of this pathogen with soybean seeds. In other countries, this pathogen has been identified as the causal agent of stem canker (Mena et al. 2020; Lopez-Cardona et al. 2021). Further research is needed to analyze the risk of this seed-associated pathogen.

Intelligent real-life key-pixel image detection system for early Arabic sign language learners.

Lack of an effective early sign language learning framework for a hard-of-hearing population can have traumatic consequences, causing social isolation and unfair treatment in workplaces. Alphabet and digit detection methods have been the basic framework for early sign language learning but are restricted by performance and accuracy, making it difficult to detect signs in real life. This article proposes an improved sign language detection method for early sign language learners based on the You Only Look Once version 8.0 (YOLOv8) algorithm, referred to as the intelligent sign language detection system (iSDS), which exploits the power of deep learning to detect sign language-distinct features. The iSDS method could overcome the false positive rates and improve the accuracy as well as the speed of sign language detection. The proposed iSDS framework for early sign language learners consists of three basic steps: (i) image pixel processing to extract features that are underrepresented in the frame, (ii) inter-dependence pixel-based feature extraction using YOLOv8, (iii) web-based signer independence validation. The proposed iSDS enables faster response times and reduces misinterpretation and inference delay time. The iSDS achieved state-of-the-art performance of over 97% for precision, recall, and F1-score with the best mAP of 87%. The proposed iSDS method has several potential applications, including continuous sign language detection systems and intelligent web-based sign recognition systems.

An intrusion detection system based on convolution neural network.

With the rapid extensive development of the Internet, users not only enjoy great convenience but also face numerous serious security problems. The increasing frequency of data breaches has made it clear that the network security situation is becoming increasingly urgent. In the realm of cybersecurity, intrusion detection plays a pivotal role in monitoring network attacks. However, the efficacy of existing solutions in detecting such intrusions remains suboptimal, perpetuating the security crisis. To address this challenge, we propose a sparse autoencoder-Bayesian optimization-convolutional neural network (SA-BO-CNN) system based on convolutional neural network (CNN). Firstly, to tackle the issue of data imbalance, we employ the SMOTE resampling function during system construction. Secondly, we enhance the system's feature extraction capabilities by incorporating SA. Finally, we leverage BO in conjunction with CNN to enhance system accuracy. Additionally, a multi-round iteration approach is adopted to further refine detection accuracy. Experimental findings demonstrate an impressive system accuracy of 98.36%. Comparative analyses underscore the superior detection rate of the SA-BO-CNN system.

Detection of fake face images using lightweight convolutional neural networks with stacking ensemble learning method.

Images and videos containing fake faces are the most common type of digital manipulation. Such content can lead to negative consequences by spreading false information. The use of machine learning algorithms to produce fake face images has made it challenging to distinguish between genuine and fake content. Face manipulations are categorized into four basic groups: entire face synthesis, face identity manipulation (deepfake), facial attribute manipulation and facial expression manipulation. The study utilized lightweight convolutional neural networks to detect fake face images generated by using entire face synthesis and generative adversarial networks. The dataset used in the training process includes 70,000 real images in the FFHQ dataset and 70,000 fake images produced with StyleGAN2 using the FFHQ dataset. 80% of the dataset was used for training and 20% for testing. Initially, the MobileNet, MobileNetV2, EfficientNetB0, and NASNetMobile convolutional neural networks were trained separately for the training process. In the training, the models were pre-trained on ImageNet and reused with transfer learning. As a result of the first trainings EfficientNetB0 algorithm reached the highest accuracy of 93.64%. The EfficientNetB0 algorithm was revised to increase its accuracy rate by adding two dense layers (256 neurons) with ReLU activation, two dropout layers, one flattening layer, one dense layer (128 neurons) with ReLU activation function, and a softmax activation function used for the classification dense layer with two nodes. As a result of this process accuracy rate of 95.48% was achieved with EfficientNetB0 algorithm. Finally, the model that achieved 95.48% accuracy was used to train MobileNet and MobileNetV2 models together using the stacking ensemble learning method, resulting in the highest accuracy rate of 96.44%.

Comparative evaluation of machine learning algorithms for phishing site detection.

The advent of Internet technologies has resulted in the proliferation of electronic trading and the use of the Internet for electronic transactions, leading to a rise in unauthorized access to sensitive user information and the depletion of resources for enterprises. As a consequence, there has been a marked increase in phishing, which is now considered one of the most common types of online theft. Phishing attacks are typically directed towards obtaining confidential information, such as login credentials for online banking platforms and sensitive systems. The primary objective of such attacks is to acquire specific personal information to either use for financial gain or commit identity theft. Recent studies have been conducted to combat phishing attacks by examining domain characteristics such as website addresses, content on websites, and combinations of both approaches for the website and its source code. However, businesses require more effective anti-phishing technologies to identify phishing URLs and safeguard their users. The present research aims to evaluate the effectiveness of eight machine learning (ML) and deep learning (DL) algorithms, including support vector machine (SVM), k-nearest neighbors (KNN), random forest (RF), Decision Tree (DT), Extreme Gradient Boosting (XGBoost), logistic regression (LR), convolutional neural network (CNN), and DL model and assess their performances in identifying phishing. This study utilizes two real datasets, Mendeley and UCI, employing performance metrics such as accuracy, precision, recall, false positive rate (FPR), and F-1 score. Notably, CNN exhibits superior accuracy, emphasizing its efficacy. Contributions include using purpose-specific datasets, meticulous feature engineering, introducing SMOTE for class imbalance, incorporating the novel CNN model, and rigorous hyperparameter tuning. The study demonstrates consistent model performance across both datasets, highlighting stability and reliability.

Outlier detection for keystroke biometric user authentication.

User authentication is a fundamental aspect of information security, requiring robust measures against identity fraud and data breaches. In the domain of keystroke dynamics research, a significant challenge lies in the reliance on imposter datasets, particularly evident in real-world scenarios where obtaining authentic imposter data is exceedingly difficult. This article presents a novel approach to keystroke dynamics-based authentication, utilizing unsupervised outlier detection techniques, notably exemplified by the histogram-based outlier score (HBOS), eliminating the necessity for imposter samples. A comprehensive evaluation, comparing HBOS with 15 alternative outlier detection methods, highlights its superior performance. This departure from traditional dependence on imposter datasets signifies a substantial advancement in keystroke dynamics research. Key innovations include the introduction of an alternative outlier detection paradigm with HBOS, increased practical applicability by reducing reliance on extensive imposter data, resolution of real-world challenges in simulating fraudulent keystrokes, and addressing critical gaps in existing authentication methodologies. Rigorous testing on Carnegie Mellon University's (CMU) keystroke biometrics dataset validates the effectiveness of the proposed approach, yielding an impressive equal error rate (EER) of 5.97%, a notable area under the ROC curve of 97.79%, and a robust accuracy (ACC) of 89.23%. This article represents a significant advancement in keystroke dynamics-based authentication, offering a reliable and efficient solution characterized by substantial improvements in accuracy and practical applicability.

A novel clustered-based detection method for shilling attack in private environments.

The topic of privacy-preserving collaborative filtering is gaining more and more attention. Nevertheless, privacy-preserving collaborative filtering techniques are vulnerable to shilling or profile injection assaults. Hence, it is crucial to identify counterfeit profiles in order to achieve total success. Various techniques have been devised to identify and prevent intrusion patterns from infiltrating the system. Nevertheless, these strategies are specifically designed for collaborative filtering algorithms that do not prioritize privacy. There is a scarcity of research on identifying shilling attacks in recommender systems that prioritize privacy. This work presents a novel technique for identifying shilling assaults in privacy-preserving collaborative filtering systems. We employ an ant colony clustering detection method to effectively identify and eliminate fake profiles that are created by six widely recognized shilling attacks on compromised data. The objective of the study is to categorize the fraudulent profiles into a specific cluster and separate this cluster from the system. Empirical experiments are conducted with actual data. The empirical findings demonstrate that the strategy derived from the study effectively eliminates fraudulent profiles in privacy-preserving collaborative filtering.

CyberEduPlatform: an educational tool to improve cybersecurity through anomaly detection with Artificial Intelligence.

Cybersecurity has become a central concern in the contemporary digital era due to the exponential increase in cyber threats. These threats, ranging from simple malware to advanced persistent attacks, put individuals and organizations at risk. This study explores the potential of artificial intelligence to detect anomalies in network traffic in a university environment. The effectiveness of automatic detection of unconventional activities was evaluated through extensive simulations and advanced artificial intelligence models. In addition, the importance of cybersecurity awareness and education is highlighted, introducing CyberEduPlatform, a tool designed to improve users' cyber awareness. The results indicate that, while AI models show high precision in detecting anomalies, complementary education and awareness play a crucial role in fortifying the first lines of defense against cyber threats. This research highlights the need for an integrated approach to cybersecurity, combining advanced technological solutions with robust educational strategies.