Neural Computing & Applications最新文献

Nowadays, quick, and accurate diagnosis of COVID-19 is a pressing need. This study presents a multimodal system to meet this need. The presented system employs a machine learning module that learns the required knowledge from the datasets collected from 930 COVID-19 patients hospitalized in Italy during the first wave of COVID-19 (March-June 2020). The dataset consists of twenty-five biomarkers from electronic health record and Chest X-ray (CXR) images. It is found that the system can diagnose low- or high-risk patients with an accuracy, sensitivity, and F1-score of 89.03%, 90.44%, and 89.03%, respectively. The system exhibits 6% higher accuracy than the systems that employ either CXR images or biomarker data. In addition, the system can calculate the mortality risk of high-risk patients using multivariate logistic regression-based nomogram scoring technique. Interested physicians can use the presented system to predict the early mortality risks of COVID-19 patients using the web-link: Covid-severity-grading-AI. In this case, a physician needs to input the following information: CXR image file, Lactate Dehydrogenase (LDH), Oxygen Saturation (O₂%), White Blood Cells Count, C-reactive protein, and Age. This way, this study contributes to the management of COVID-19 patients by predicting early mortality risk.

Supplementary information: The online version contains supplementary material available at 10.1007/s00521-023-08606-w.

Covid-19 is a very dangerous disease as a result of the rapid and unprecedented spread of any previous disease. It is truly a crisis that threatens the world since its first appearance in December 2019 until our time. Due to the lack of a vaccine that has proved sufficiently effective so far, the rapid and more accurate diagnosis of this disease is extremely necessary to enable the medical staff to identify infected cases and isolate them from the rest to prevent further loss of life. In this paper, Covid-19 diagnostic strategy (CDS) as a new classification strategy that consists of two basic phases: Feature selection phase (FSP) and diagnosis phase (DP) has been introduced. During the first phase called FSP, the best set of features in laboratory test findings for Covid-19 patients will be selected using enhanced gray wolf optimization (EGWO). EGWO combines both types of selection techniques called wrapper and filter. Accordingly, EGWO includes two stages called filter stage (FS) and wrapper stage (WS). While FS uses many different filter methods, WS uses a wrapper method called binary gray wolf optimization (BGWO). The second phase called DP aims to give fast and more accurate diagnosis using a hybrid diagnosis methodology (HDM) based on the selected features from FSP. In fact, the HDM consists of two phases called weighting patient phase (WP²) and diagnostic patient phase (DP²). WP² aims to calculate the belonging degree of each patient in the testing dataset to class category using naïve Bayes (NB) as a weight method. On the other hand, K-nearest neighbor (KNN) will be used in DP² based on the weights of patients in the testing dataset as a new training dataset to give rapid and more accurate detection. The suggested CDS outperforms other strategies according to accuracy, precision, recall (or sensitivity) and F-measure calculations that are equal to 99%, 88%, 90% and 91%, respectively, as showed in experimental results.

A novel multi-objective Coronavirus disease optimization algorithm (MOCOVIDOA) is presented to solve global optimization problems with up to three objective functions. This algorithm used an archive to store non-dominated POSs during the optimization process. Then, a roulette wheel selection mechanism selects the effective archived solutions by simulating the frameshifting technique Coronavirus particles use for replication. We evaluated the efficiency by solving twenty-seven multi-objective (21 benchmarks & 6 real-world engineering design) problems, where the results are compared against five common multi-objective metaheuristics. The comparison uses six evaluation metrics, including IGD, GD, MS, SP, HV, and delta p ($\Delta P$). The obtained results and the Wilcoxon rank-sum test show the superiority of this novel algorithm over the existing algorithms and reveal its applicability in solving multi-objective problems.

During the past three years, the coronavirus disease 2019 (COVID-19) has swept the world. The rapid and accurate recognition of covid-19 pneumonia are ,therefore, of great importance. To handle this problem, we propose a new pipeline of deep learning framework for diagnosing COVID-19 pneumonia via chest X-ray images from normal, COVID-19, and other pneumonia patients. In detail, the self-trained YOLO-v4 network was first used to locate and segment the thoracic region, and the output images were scaled to the same size. Subsequently, the pre-trained convolutional neural network was adopted to extract the features of X-ray images from 13 convolutional layers, which were fused with the original image to form a 14-dimensional image matrix. It was then put into three parallel pyramid multi-layer perceptron (MLP)-Mixer modules for comprehensive feature extraction through spatial fusion and channel fusion based on different scales so as to grasp more extensive feature correlation. Finally, by combining all image features from the 14-channel output, the classification task was achieved using two fully connected layers as well as Softmax classifier for classification. Extensive simulations based on a total of 4099 chest X-ray images were conducted to verify the effectiveness of the proposed method. Experimental results indicated that our proposed method can achieve the best performance in almost all cases, which is good for auxiliary diagnosis of COVID-19 and has great clinical application potential.

Time-series prediction and imputation receive lots of attention in academic and industrial areas. Machine learning methods have been developed for specific time-series scenarios; however, it is difficult to evaluate the effectiveness of a certain method on other new cases. In the perspective of frequency features, a comprehensive benchmark for time-series prediction is designed for fair evaluation. A prediction problem generation process, composed of the finite impulse response filter-based approach and problem setting module, is adopted to generate the NCAA2022 dataset, which includes 16 prediction problems. To reduce the computational burden, the filter parameters matrix is divided into sub-matrices. The discrete Fourier transform is introduced to analyze the frequency distribution of transformed results. In addition, a baseline experiment further reflects the benchmarking capability of NCAA2022 dataset.

Uniform Resource Locator (URL) is a unique identifier composed of protocol and domain name used to locate and retrieve a resource on the Internet. Like any Internet service, URLs (also called websites) are vulnerable to compromise by attackers to develop Malicious URLs that can exploit/devastate the user's information and resources. Malicious URLs are usually designed with the intention of promoting cyber-attacks such as spam, phishing, malware, and defacement. These websites usually require action on the user's side and can reach users across emails, text messages, pop-ups, or devious advertisements. They have a potential impact that can reach, in some cases, to compromise the machine or network of the user, especially those arriving by email. Therefore, developing systems to detect malicious URLs is of great interest nowadays. This paper proposes a high-performance machine learning-based detection system to identify Malicious URLs. The proposed system provides two layers of detection. Firstly, we identify the URLs as either benign or malware using a binary classifier. Secondly, we classify the URL classes based on their feature into five classes: benign, spam, phishing, malware, and defacement. Specifically, we report on four ensemble learning approaches, viz. the ensemble of bagging trees (En_Bag) approach, the ensemble of k-nearest neighbor (En_kNN) approach, and the ensemble of boosted decision trees (En_Bos) approach, and the ensemble of subspace discriminator (En_Dsc) approach. The developed approaches have been evaluated on an inclusive and contemporary dataset for uniform resource locators (ISCX-URL2016). ISCX-URL2016 provides a lightweight dataset for detecting and categorizing malicious URLs according to their attack type and lexical analysis. Conventional machine learning evaluation measurements are used to evaluate the detection accuracy, precision, recall, F Score, and detection time. Our experiential assessment indicates that the ensemble of bagging trees (En_Bag) approach provides better performance rates than other ensemble methods. Alternatively, the ensemble of the k-nearest neighbor (En_kNN) approach provides the highest inference speed. We also contrast our En_Bag model with state-of-the-art solutions and show its superiority in binary classification and multi-classification with accuracy rates of 99.3% and 97.92%, respectively.

The majority of Virtual Private Networks (VPNs) fail when it comes to protecting our privacy. If we are using a VPN to protect our online privacy, many of the well-known VPNs are not secure to use. When examined closely, VPNs can appear to be perfect on the surface but still be a complete privacy and security disaster. Some VPNs will steal our bandwidth, infect our computers with malware, install secret tracking libraries on our devices, steal our personal data, and leave our data exposed to third parties. Generally, Android users should be cautious when installing any VPN software on their devices. As a result, it is important to identify malicious VPNs before downloading and installing them on our Android devices. This paper provides an optimised deep learning neural network for identifying fake VPNs, and VPNs infected by malware based on the permissions of the apps, as well as a novel dataset of malicious and benign Android VPNs. Experimental results indicate that our proposed classifier identifies malicious VPNs with high accuracy, while it outperforms other standard classifiers in terms of evaluation metrics such as accuracy, precision, and recall.

Digital data security has become an exigent area of research due to a huge amount of data availability at present time. Some of the fields like medical imaging and medical data sharing over communication platforms require high security against counterfeit access, manipulation and other processing operations. It is essential because the changed/manipulated data may lead to erroneous judgment by medical experts and can negatively influence the human's heath. This work offers a blind and robust medical image watermarking framework using deep neural network to provide effective security solutions for medical images. During watermarking, the region of interest (ROI) data of the original image is preserved by employing the LZW (Lampel-Ziv-Welch) compression algorithm. Subsequently the robust watermark is inserted into the original image using IWT (integer wavelet transform) based embedding approach. Next, the SHA-256 algorithm-based hash keys are generated for ROI and RONI (region of non-interest) regions. The fragile watermark is then prepared by ROI recovery data and the hash keys. Further, the LSB replacement-based insertion mechanism is utilized to embed the fragile watermark into RONI embedding region of robust watermarked image. A deep neural network-based framework is used to perform robust watermark extraction for efficient results with less computational time. Simulation results verify that the scheme has significant imperceptibility, efficient robust watermark extraction, correct authentication and completely reversible nature for ROI recovery. The relative investigation with existing schemes confirms the dominance of the proposed work over already existing work.