Low Detection Rate Research Articles

EGMA: Ensemble Learning-Based Hybrid Model Approach for Spam Detection

Spam messages have emerged as a significant issue in digital communication, adversely affecting users’ mental health, personal safety, and network resources. Traditional spam detection methods often suffer from low detection rates and high false positives, underscoring the need for more effective solutions. This paper proposes the EGMA model, an ensemble learning-based hybrid approach for spam detection in SMS messages, which integrates gated recurrent unit (GRU), multilayer perceptron (MLP), and hybrid autoencoder models utilizing a majority voting algorithm. The EGMA model enhances performance by incorporating additional statistical features extracted from message content and employing text vectorization techniques, such as Term Frequency–Inverse Document Frequency (TF-IDF) and CountVectorizer. The proposed model achieved impressive classification accuracies of 99.28% on the SMS Spam Collection dataset, 99.24% on the Email Spam dataset, 99.00% on the Enron-Spam dataset, 98.71% on the Super SMS dataset, and 95.09% on UtkMl’s Twitter Spam dataset. These results demonstrate that the EGMA model outperforms individual models and existing methods in the literature, providing a robust solution for enhancing spam detection performance and effectively mitigating the threats that spam messages pose in digital communication.

Observing nociceptive detection thresholds and evoked potentials in diabetic patients with and without painful neuropathy

Aim: Diabetic polyneuropathy is the most described complication in patients with diabetes mellitus. A significant percentage of these patients experience disabling neuropathic pain (painful diabetic polyneuropathy). Small nerve fibers are primarily responsible for peripheral nociception, but objectively assessing its function is challenging. The primary objective of this study was to explore the task execution and outcomes of intra-epidermal electrical stimulation technique that combines nociceptive detection thresholds (NDT) and evoked potentials (EPs) in patients with diabetes. We compared the results of diabetic patients, both with and without painful diabetic polyneuropathy, with those of healthy controls to explore potential clinically relevant information. Methods: The NDT-EP method was applied to 38 patients with diabetes (18 with and 20 without chronic painful neuropathy) and 38 age- and sex-matched healthy controls. Individual mean NDTs, psychometric slopes, EP amplitudes, and the effect of the stimuli on EP amplitudes were analyzed and compared between groups using linear regression. Results: The findings revealed significantly lower detection rates, higher NDTs, and lower psychometric slopes in patients with painful diabetic polyneuropathy than in healthy controls. Both patient groups significantly exhibited lower mean EP amplitudes than healthy controls, which were not linked to pulse amplitudes but influenced by stimulus detection. Conclusions: This study showed altered NDT-EP outcomes in patients with painful diabetic polyneuropathy. Whereas the task execution, NDTs, and psychometric slopes may provide valuable insights into small fiber dysfunction, pulse amplitudes seemed not differently encoded in neurophysiological responses to intra-epidermal electrical stimulation near the detection threshold compared to controls. Future studies should investigate whether the altered NDT-EP outcomes could quantify small fiber dysfunction in patients with diabetes mellitus. We recommend further exploration of NDT-EP measures in other patient groups with nociceptive dysfunction.

Forensic STR and SNP Genotyping of Formalin-Fixed Skeleton Samples With Illumina's ForenSeq System.

Formalin fixatives are widely used in forensics to preserve tissues. However, extracting high-quality genomic DNA from formalin-fixed samples is challenging. Traditional short tandem repeat (STR) analysis using capillary electrophoresis (CE) for forensic DNA typing frequently results in failure. Massively parallel sequencing (MPS) can handle many samples and thousands of genetic markers, usually single-nucleotide polymorphisms (SNPs) and STRs, in a single test. Thus, it is useful for assessing highly deteriorated forensic evidence. Few studies have examined the effectiveness of STRs and SNPs genotyping of formalin-fixed skeletons using MPS. In this study, 55 skeletal samples from 5 individuals that were treated under different formalin fixation times (5-75 days) were examined and sequenced using the ForenSeq DNA Signature Prep Kit on the Illumina MiSeq FGX platform. The results showed that as the duration of formalin fixation increased, the detection rates of STRs and SNPs gradually decreased. After 75 days of fixation, the average detection rates for STRs and SNPs were 4% and 10%, respectively. The cumulative discrimination power (CDP) of individual identification SNPs (iiSNPs) was >0.9999 on the 45th day. However, the CDP of STRs was 0.9930 on the 22nd day. Low detection rates were observed for six STRs (D1S1656, PentaE, D22S1045, PentaD, DX8378 and DX10103) and five SNPs (rs2920816, rs354439, rs1736442, rs338882 and rs1031825). In conclusion, DNA extracted from formalin-fixed skeletons decomposes rapidly over time, and MPS technology can be a useful tool for detecting forensic genetic markers in such samples.

Spatial–Temporal Patterns in the Enteric Pathogen Contamination of Soil in the Public Environments of Low- and Middle-Income Neighborhoods in Nairobi, Kenya

Public spaces in countries with limited societal development can be contaminated with feces containing pathogenic microbes from animals and people. Data on contamination levels, spatial distribution, and the diversity of enteric pathogens in the public settings of low- and middle-income neighborhoods are crucial for devising strategies that minimize the enteric infection burden. The objective of this study was to compare spatial–temporal differences in the detection rate and diversity of enteric pathogens in the public spaces of low- and middle-income neighborhoods of Nairobi, Kenya. TaqMan array card (TAC) molecular assays were employed to analyze soil samples for 19 enteropathogens, along with a selective bacterial culture for pathogenic Enterobacteriaceae. An observational assessment was conducted during every site visit to document the hygienic infrastructure and sanitation conditions at the sites. We detected at least one pathogen in 79% (127/160) and ≥2 pathogens in 67.5% (108/160) of the soil samples tested. The four most frequently detected pathogens were EAEC (67.5%), ETEC (59%), EPEC (57.5%), and STEC (31%). The detection rate (91% vs. 66%) and mean number of enteric pathogens (5 vs. 4.7) were higher in low-income Kibera than in middle-income Jericho. The more extensive spatial distribution of pathogens in Kibera resulted in increases in the detection of different enteric pathogens from within-site (area < 50 m2) and across-site (across-neighborhood) movements compared to Jericho. The pathogen detection rates fluctuated seasonally in Jericho but remained at sustained high levels in Kibera. While better neighborhood conditions were linked with lower pathogen detection rates, pathogenic E. coli remained prevalent in the public environment across both neighborhoods. Future studies should focus on identifying how the sources of pathogen contamination are modified by improved environmental sanitation and hygiene and the role of these contaminated public environments in enteric infections in children.

Video SAR Moving Target Detection Method Based on Machine Vision

When detecting moving targets in video SAR, it is impossible to accurately obtain the moving targets in dynamic images. In order to solve the problems of low detection integrity and poor effectiveness, a video SAR moving target detection method based on machine vision is proposed. Firstly, the moving target pixel points are matched by the mixed Gaussian model to obtain the target area of the video SAR moving image; secondly, the HSV model and reflectivity algorithm are used to remove the shadows caused by the moving target; finally, the processed target area is input into the YOLO algorithm to complete the final video SAR moving target detection. Experimental results show that the proposed algorithm has high detection integrity, high detection rate, low false detection rate, and better detection effectiveness.

Temperature Measurement Timings and the Fever Detection Rate After Gastrointestinal Surgery: Retrospective Cross-Sectional Study.

Postoperative fever frequently indicates surgical complications and is commonly used to evaluate the efficacy of interventions against surgical stress. However, the presence of circadian rhythms in body temperature may compromise the accurate detection of fever. This study aimed to investigate the detection rate of fever under intermittent measurement. We retrospectively reviewed the clinical records of patients who underwent nonemergency gastrointestinal surgery between November 2020 and April 2021. Patients' temperature data were continuously collected every 4 seconds using a wireless axillary thermometer, and fever was defined as a temperature exceeding 38 °C within a day. To simulate intermittent measurement in clinical practice, the body temperature at each hour was selected from the continuously collected temperature dataset. Considering that temperatures are measured multiple times per day, all possible measurement plans using intermittent measurement were composed by combining 1-24 time points from the 24-hour daily cycle. Fever was clinically diagnosed based on the temperature readings at the selected time points per day. The fever detection rates for each plan, with varying measurement times, were listed and ranked. Based on the temperature data continuously collected by the thermometer, fever occurred in 60 (40.8%) of the 147 included patients within 3 days after surgery. Of the measurement plans that included 1-24 measurements daily, the fever detection rates ranged from 3.3% (2/60) to 85% (51/60). The highest detection rates and corresponding timings for measurement plans with 1, 2, 3, and 4 measurements daily were 38.3% (23/60; at 8 PM), 56.7% (34/60; at 3 AM and 7 or 8 PM), 65% (39/60; at 3 AM, 8 PM, and 10 or 11 PM), and 70% (42/60; at 12 AM, 3 AM, 8 PM, and 11 PM), respectively; and the lowest detection rates were 3.3% (2/60), 6.7% (4/60), 6.7% (4/60), and 8.3% (5/60), respectively. Although fever within 3 days after surgery was not correlated with an increased incidence of postoperative complications (5/60, 8.3% vs 6/87, 6.9%; P=.76), it was correlated with a longer hospital stay (median 7, IQR 6-9 days vs median 6, IQR 5-7 days; P<.001). The fever detection rate of the intermittent approach is determined by the timing and frequency of measurement. Measuring at randomly selected time points can miss many fever events after gastrointestinal surgery. However, we can improve the fever detection rate by optimizing the timing and frequency of measurement.

Network anomaly detection using Deep Autoencoder and parallel Artificial Bee Colony algorithm-trained neural network

Cyberattacks are increasingly becoming more complex, which makes intrusion detection extremely difficult. Several intrusion detection approaches have been developed in the literature and utilized to tackle computer security intrusions. Implementing machine learning and deep learning models for network intrusion detection has been a topic of active research in cybersecurity. In this study, artificial neural networks (ANNs), a type of machine learning algorithm, are employed to determine optimal network weight sets during the training phase. Conventional training algorithms, such as back-propagation, may encounter challenges in optimization due to being entrapped within local minima during the iterative optimization process; global search strategies can be slow at locating global minima, and they may suffer from a low detection rate. In the ANN training, the Artificial Bee Colony (ABC) algorithm enables the avoidance of local minimum solutions by conducting a high-performance search in the solution space but it needs some modifications. To address these challenges, this work suggests a Deep Autoencoder (DAE)-based, vectorized, and parallelized ABC algorithm for training feed-forward artificial neural networks, which is tested on the UNSW-NB15 and NF-UNSW-NB15-v2 datasets. Our experimental results demonstrate that the proposed DAE-based parallel ABC-ANN outperforms existing metaheuristics, showing notable improvements in network intrusion detection. The experimental results reveal a notable improvement in network intrusion detection through this proposed approach, exhibiting an increase in detection rate (DR) by 0.76 to 0.81 and a reduction in false alarm rate (FAR) by 0.016 to 0.005 compared to the ANN-BP algorithm on the UNSW-NB15 dataset. Furthermore, there is a reduction in FAR by 0.006 to 0.0003 compared to the ANN-BP algorithm on the NF-UNSW-NB15-v2 dataset. These findings underscore the effectiveness of our proposed approach in enhancing network security against network intrusions.

The Utility of an Human Papillomavirus Genotype Assay for Cancer Screening in Self-Collected Urine and Vaginal Samples from Japanese Women

Objectives: The high incidence of invasive cervical cancer among those who have not undergone cancer screening is a serious problem. This study aimed to investigate the utility of human papillomavirus (HPV) test results from self-collected urine and vaginal samples as screening tools. Design: The study was conducted in two steps. First, the appropriate storage container, temperature, and time until urine HPV assay performance were verified. Second, the results of spot urine testing under those conditions and of gynecologist-collected cervical and self-collected vaginal samples were compared to verify the feasibility of using the BD Onclarity® HPV assay for individuals with abnormal cervical cytology. Participants/Materials, Setting, Methods: The participants were 121 women with abnormal cervical cytology. Self-collected urine and vaginal samples, along with gynecologist-collected cervical samples, were tested for HPV using the BD Onclarity® HPV assay. The optimal conditions for urine sample storage were identified by comparing the HPV detection rates under various conditions. Results: Urine stored in a BD Probe Tec™ (QxUPT) for less than 72 h at room temperature was found to have the highest HPV positivity rate. Under these conditions, the detection rates of HPV in urine, cervical, and vaginal samples were examined. HPV type 16 was detected in 41.7% of the cervical samples, type 18 in 10%, and types 31 and 52 in 12.6% each. The concordance rate for HPV testing between clinician-collected cervical and urine samples was 63.9% (kappa: 0.34; 95% CI: 0.21–0.47), and that between clinician-collected cervical and self-collected vaginal samples was 77.8% (kappa: 0.68; 95% CI: 0.53–0.83), indicating good concordance. In a population with an HPV-related lesion/tumor prevalence of approximately 70%, the sensitivity of HPV testing was 82.7% for the cervix, 46.4% for urine, and 75.7% for vaginal samples. Limitations: The primary limitation is the lower detection rate of HPV in spot urine samples than in other sample types, indicating room for methodological improvement. The study’s findings are based on a specific population, which may limit generalizability. Conclusions: We investigated the optimal self-collected urine-to-testing time and temperature. Self-collected vaginal and urine HPV tests show moderate-high concordance with clinician-collected cervical HPV tests, suggesting their potential utility for women who do not undergo regular cancer screening. However, the sensitivity was not high in spot urine. Therefore, further large-scale studies are needed to verify these findings and optimize testing methods to encourage broader participation in cancer screening programs.

One-Pot Assay for Rapid Detection of Stenotrophomonas maltophilia by RPA-CRISPR/Cas12a.

Stenotrophomonas maltophilia (S. maltophilia, SMA) is a common opportunistic pathogen that poses a serious threat to the food industry and human health. Traditional detection methods for SMA are time-consuming, have low detection rates, require complex and expensive equipment and professional technical personnel for operation, and are unsuitable for on-site detection. Therefore, establishing an efficient on-site detection method has great significance in formulating appropriate treatment strategies and ensuring food safety. In the present study, a rapid one-pot detection method was established for SMA using a combination of Recombinase Polymerase Amplification (RPA) and CRISPR/Cas12a, referred to as ORCas12a-SMA (one-pot RPA-CRISPR/Cas12a platform). In the ORCas12a-SMA detection method, all components were added into a single tube simultaneously to achieve one-pot detection and address the problems of nucleic acid cross-contamination and reduced sensitivity caused by frequent cap opening during stepwise detection. The ORCas12a-SMA method could detect at least 3 × 10° copies·μL-1 of SMA genomic DNA within 30 min at 37 °C. Additionally, this method exhibited sensitivity compared to the typical two-step RPA-CRISPR/Cas12a method. Overall, the ORCas12a-SMA detection offered the advantages of rapidity, simplicity, high sensitivity and specificity, and decreased need for complex large-scale instrumentation. This assay is the first application of the one-pot platform based on the combination of RPA and CRISPR/Cas12a in SMA detection and is highly suitable for point-of-care testing. It helps reduce losses in the food industry and provides assistance in formulating timely and appropriate antimicrobial treatment plans.

A robust detection scheme against selective forwarding attacks in wireless sensor networks under variable harsh environments

Event-driven wireless sensor networks (EWSNs) often operate in harsh environments and are susceptible to selective forwarding attacks, where malicious nodes selectively drop packets. These attacks are difficult to detect due to similarities with normal node behavior under adverse conditions. To address this, we propose a detection scheme combining Variable Autoencoder (VAE) and Gaussian Mixture Model (GMM) with an Autoencoder (AE). The scheme extracts features from the node's single round forwarding rate (SFR), with VAE generating latent variables and GMM clustering them to differentiate behaviors. The AE model, trained on data representing normal behavior, uses reconstruction error to identify malicious nodes. Simulation results show that our scheme improves network throughput and achieves low False Detection Rate (FDR) and Missed Detection Rate (MDR), with both rates as low as 0.5% in normal conditions and below 2% in mobile harsh environments. With a computational complexity of O(n), this method is suitable for resource-constrained environments, demonstrating superior robustness compared to existing approaches.

Modelling multiplex testing for outbreak control

During the SARS-CoV-2 pandemic, polymerase chain reaction (PCR) and lateral flow device (LFD) tests were frequently deployed to detect the presence of SARS-CoV-2. Many of these tests were singleplex, and only tested for the presence of a single pathogen. Multiplex tests can test for the presence of several pathogens using only a single swab, which can allow for: surveillance of more pathogens, targeting of antiviral interventions, a reduced burden of testing, and lower costs. Test sensitivity however, particularly in LFD tests, is highly conditional on the viral concentration dynamics of individuals. To inform the use of multiplex testing in outbreak detection it is therefore necessary to investigate the interactions between outbreak detection strategies and the differing viral concentration trajectories of key pathogens. Viral concentration trajectories are estimated for SARS-CoV-2, and Influenza A/B. Testing strategies for the first five symptomatic cases in an outbreak are then simulated and used to evaluate key performance indicators. Strategies that use a combination of multiplex LFD and PCR tests achieve; high levels of detection, detect outbreaks rapidly, and have the lowest burden of testing across multiple pathogens. Influenza B was estimated to have lower rates of detection due to its modelled viral concentration dynamics. Data availability statementThe SARS-CoV-2 viral load trajectory data may be accessed by contacting Killingley et. al. The influenza A/B viral load trajectory may be accessed at https://github.com/bcowling/pediatric-vaccine-trial/tree/master/data.

Intrusion Detection System Using Chaotic Walrus Optimization-based Convolutional Echo State Networks for IoT-assisted Wireless Sensor Networks

Wireless Sensor Networks (WSN) based Internet of Things (IoT) networks have achieved greater research interest due to their multi-purpose data collection and transmission over different geographical locations. However, the fabrication of different cyber-attacks in these networks has been a severe concern for applications such as remote healthcare, military communications, etc. These attacks question the integrity and security of WSN-IOT networks for such applications, and the traditional Intrusion Detection Systems (IDS) have shown increased false alarm rates because of their substantial processing overhead, resource constraints, and low detection rates. This paper presents an intelligent IDS model using Chaotic Walrus Optimization-based Convolutional Echo State Networks (CWO-CESN) to solve existing problems. It increases the detection accuracy of different attacks. In this CWO-CESN-based IDS model, the data are gathered from the sensor/IoT nodes and are pre-processed. Then, the proposed CWO-CESN learns the features from these data and classifies them into attacks and standard data classes. This proposed CWO-CESN is a hybrid classifier model that integrates Convolutional Neural Networks (CNN) and Echo State Networks (ESN) as a single model and employs Chaotic map-based population initialized Walrus Optimizer for optimizing the hyperparameters. Validated on benchmark datasets, the proposed CWO-CESN-based IDS model attained accuracies of 99%, 99.5%, and 99.8% for the detection of different attacks for NSL-KDD, WSN-DS and IoT-23 datasets, respectively, and ensured secured and reliable application in significant fields.

Deep Packet Inspection Model Based on Support Vector Machine for Anomaly Detection in Local Area Networks

Deep packet inspection is a network security solution that identifies and flags anomalous network traffic patterns in a local network environment. Traditional signature-based techniques for intrusion detection are limited in identifying different attacks or completely new kinds, which makes them unsuitable in some situations. In addition, most previous methods for anomaly detection have low detection rate and high false alarm. In this study, a deep packet inspection model based on support vector machine (SVM) for anomaly detection in local area networks was proposed. The proposed method combined the SelectKBest method and SVM for the categorization of anomaly in a local network environment. Results showed that the proposed method outperformed other related machine learning methods with accuracy, precision, recall, and F1-score of 94.81%, 94.03%, 94.13%, and 94.0799%, respectively. The accuracy result shows that most network traffic can be correctly identified by the SVM using the SelectKBest approach, with minimal false positives or negatives.

Survey effort and targeted landbird community metrics at Indiana lowland forest restoration sites.

Many sampling and analytical methods can estimate the abundance, distributions, and diversity of birds and other wildlife. However, challenges with sample size and analytical capacity can make these methods difficult to implement for resource-limited monitoring programs. To apprise efficient and attainable sampling designs for landbird monitoring programs with limited observational data, we used breeding season bird point survey data collected in 2016 at four forest restoration sites in Indiana, USA. We evaluated three subsets of observed species richness, total possibly breeding landbirds, Partners in Flight Regional Conservation Concern (PIF RCC) landbirds, and interior forest specialists (IFSs). Simulated surveys based on field data were used to conduct Bayesian Michaelis-Menten curve analyses estimating observed species as a function of sampling effort. On comparing simulated survey sets with multiple habitat types versus those with one habitat, we found that those with multiple habitat types had estimated 39%-83% greater observed PIF RCC species richness and required 41%-55% fewer visits per point to observe an equivalent proportion of PIF RCC species. Even with multiple habitats in a survey set, the number of visits per point required to detect 50% of observable species was 30% higher for PIF RCC species than for total breeding landbird species. Low detection rates of IFS species at two field sites made precise estimation of required effort to observe these species difficult. However, qualitatively, we found that only sites containing mature forest fragments had detections of several bird species designated as high-confidence IFS species. Our results suggest that deriving specialized species diversity metrics from point survey data can add value when interpreting those data. Additionally, designing studies to collect these metrics may require explicitly planning to visit multiple habitat types at a monitoring site and increasing the number of visits per survey point. Integr Environ Assess Manag 2024;20:1954-1968. © 2024 The Author(s). Integrated Environmental Assessment and Management published by Wiley Periodicals LLC on behalf of Society of Environmental Toxicology & Chemistry (SETAC). This article has been contributed to by U.S. Government employees and their work is in the public domain in the USA.

Research on Welding Defect Classification Methods Based on Convolutional Neural Networks

To address the issues of low detection and identification rates and inaccurate classification of welding defects, this paper proposes a welding defect classification method based on convolutional neural networks. By modifying the ResNet101 backbone network and incorporating a multi-head attention mechanism (MHA), the method enhances the focus on regions of interest in welding defect images, thereby improving the ability to extract defect features. This approach reduces the chances of missed detections and false positives. By analyzing the effects of different optimizers and dynamic factors on recall (M) and precision (N), the optimal optimizer and dynamic factor are identified. The algorithm is validated on the JPEGWD dataset. Experimental results show that the recall (M) reaches 0.9634, representing a 4.88% improvement, and the precision (N) reaches 0.9875, representing a 3.81% improvement. The mean average precision (mAP) is 0.9817, reflecting a 5.98% improvement. These results effectively enhance the recognition rate of welding defects and reduce the probability of missed detections and false positives.

Tire Wear Chemicals in the Urban Atmosphere: Significant Contributions of Tire Wear Particles to PM2.5.

Tire wear particles (TWPs) containing tire wear chemicals (TWCs) are of global concern due to their large emissions and potential toxicity. However, TWP contributions to urban fine particles are poorly understood. Here, 72 paired gas-phase and PM2.5 samples were collected in the urban air of the Pearl River Delta, China. The concentrations of 54 compounds were determined, and 28 TWCs were detected with total concentrations of 3130-317,000 pg/m3. Most p-phenylenediamines (PPDs) were unstable in solvent, likely leading to their low detection rates. The TWCs were mainly (73 ± 26%) in the gas phase. 2-OH-benzothiazole contributed 82 ± 21% of the gas-phase TWCs and benzothiazole-2-sulfonic acid contributed 74 ± 18% of the TWCs in PM2.5. Guangzhou and Foshan were "hotspots" for atmospheric TWCs. Most TWC concentrations significantly correlated with the road length nearby. More particulate TWCs were observed than model predictions, probably due to the impacts of nonexchangeable portion and sampling artifacts. Source apportionment combined with characteristic molecular markers indicated that TWPs contributed 13 ± 7% of urban PM2.5. Our study demonstrates that TWPs are important contributors to urban air pollution that could pose risks to humans. There is an urgent need to develop strategies to decrease TWP emissions, along with broader urban air quality improvement strategies.

A lightweight intrusion detection algorithm for IoT based on data purification and a separable convolution improved CNN

With the rapid development of the IoT (Internet of Things), the network data present the characteristics of large volume and high dimension. Convolutional neural networks (CNNs) have become one of the most important intrusion detection methods due to their advantages in processing high-dimensional data. The conventional intrusion detection model based on CNN lacks an effective data purification means in the process of converting unstructured data into image data, and too many parameters are generated due to the complex structure of the model in the training process, leading to the problems of high time complexity and low detection rate of the model, which limits the application of the CNN in intrusion detection of the IoT. First, based on the principle of liquid molecular distillation, a data purification algorithm (DPA) for unstructured data is proposed in this paper, which reduces the "redundant" data generated in the process of converting unstructured data to image data. Second, based on the rigid-motion convolution principle of a separable wavelet, separable convolution is used to improve the CNN structure, and then a lightweight detection algorithm LSCNN (lightweight CNN based on separable convolution) is developed to reduce the number of parameters in the network structure and improve the time efficiency and accuracy of the algorithm. The experimental results on real intrusion detection datasets show that the LSCNN trained on DPA purified data has higher time efficiency and detection accuracy than the conventional CNN, and compared with the conventional machine learning algorithm, it has higher accuracy.

Molecular diagnostic yield of exome sequencing in a Chinese cohort of 512 fetuses with anomalies

BackgroundCurrently, whole exome sequencing has been performed as a helpful complement in the prenatal setting in case of fetal anomalies. However, data on its clinical utility remain limited in practice. Herein, we reported our data of fetal exome sequencing in a cohort of 512 trios to evaluate its diagnostic yield.MethodsIn this retrospective cohort study, the couples performing prenatal exome sequencing were enrolled. Fetal phenotype was classified according to ultrasound and magnetic resonance imaging findings. Genetic variants were analyzed based on a phenotype-driven followed by genotype-driven approach in all trios.ResultsA total of 97 diagnostic variants in 65 genes were identified in 69 fetuses, with an average detection rate of 13.48%. Skeletal and renal system were the most frequently affected organs referred for whole exome sequencing, with the highest diagnostic rates. Among them, short femur and kidney cyst were the most common phenotype. Fetal growth restriction was the most frequently observed phenotype with a low detection rate (4.3%). Exome sequencing had limited value in isolated increased nuchal translucency and chest anomalies.ConclusionsThis study provides our data on the detection rate of whole exome sequencing in fetal anomalies in a large cohort. It contributes to the expanding of phenotypic and genotypic spectrum.

Research on Network Intrusion Detection Technology Based on Artificial Intelligence

As network intrusion becomes more and more complicated and network characteristics become more and more diverse, the traditional network intrusion detection technology has some problems such as high false alarm rate, poor adaptability and low detection rate. The article provides an in-depth analysis of network intrusion detection techniques based on artificial intelligence to better solve these problems. In recent years, we have been witnessing great progress in image recognition, speech recognition, and natural language processing. The powerful advantage of artificial intelligence in processing large-scale and complex data provides new ideas for the processing of multi-attribute intrusion data. The introduction of AI technology in network intrusion detection can effectively improve the detection accuracy and reduce the false alarm rate and omission rate. Based on this, the article analyses the artificial intelligence network intrusion detection technology and discusses the network security management based on artificial intelligence for the reference of related personnel.

An adaptive detection framework based on artificial immune for IoT intrusion detection system

Given the continual evolution of new network attack methodologies, defenders face the imperative of constantly upgrading security defenses. Current security technologies, albeit effective against known threats, often fall short in handling the intricacies of diverse and novel attacks. Artificial immunity-based network anomaly detection offers a promising avenue by dynamically adapting to evolving threats. However, prevailing algorithms in this domain suffer from low detection rates, limited adaptability, and extended detector generation times. This study aims to tackle these challenges by introducing a high-efficiency network anomaly detection framework, emphasizing both high-dimensional feature selection and adaptive detector generation. Our approach begins with an enhanced dual-module hybrid high-dimensional feature selection method, leveraging evolutionary principles. Furthermore, we introduce a self-sample clustering algorithm based on fuzzy clustering during the tolerance stage, enhancing detector tolerance efficiency. Additionally, an adaptive detector generation scheme is devised. It divides the non-boundary sub-population based on non-self differences and evolution, while employing the red fox optimization algorithm in the boundary region. This adaptive approach dynamically adjusts detector positions and radii to derive optimal detectors. Through comprehensive validation using well-established IoT and network anomaly datasets, our proposed artificial immunity-based IoT intrusion detection framework exhibits superior performance. It achieves higher classification accuracy and lower error rates compared to current state-of-the-art machine learning and artificial immunity algorithms.