Accurate detection of COVID-19 is of immense importance to help physicians intervene with appropriate treatments. Although RT-PCR is routinely used for COVID-19 detection, it is expensive, takes a long time, and is prone to inaccurate results. Currently, medical imaging-based detection systems have been explored as an alternative for more accurate diagnosis. In this work, we propose a multi-level diagnostic framework for the accurate detection of COVID-19 using X-ray scans based on transfer learning. The developed framework consists of three stages, beginning with a pre-processing step to remove noise effects and image resizing followed by a deep learning architecture utilizing an Xception pre-trained model for feature extraction from the pre-processed image. Our design utilizes a global average pooling (GAP) layer for avoiding over-fitting, and an activation layer is added in order to reduce the losses. Final classification is achieved using a softmax layer. The system is evaluated using different activation functions and thresholds with different optimizers. We used a benchmark dataset from the kaggle website. The proposed model has been evaluated on 7395 images that consist of 3 classes (COVID-19, normal and pneumonia). Additionally, we compared our framework with the traditional pre-trained deep learning models and with other literature studies. Our evaluation using various metrics showed that our framework achieved a high test accuracy of 99.3% with a minimum loss of 0.02 using the LeakyReLU activation function at a threshold equal to 0.1 with the RMSprop optimizer. Additionally, we achieved a sensitivity and specificity of 99 and F1-Score of 99.3% with only 10 epochs and a ${10}^{-4}$ learning rate.

Detection and classification of brain tumors are of formidable importance in neuroscience. Deep learning (DL), specifically convolution neural networks (CNN), has demonstrated breakthroughs in the field of brain image analysis and brain tumors classification. This work proposes a novel CNN based model for brain tumor classification. Our pipeline starts with prepossessing and data augmentation techniques. Then, a CNN classification step is developed and utilizes ResNet50 architecture as its core. Particularly, our design modified the ResNet50 output with a global average pooling (GAP) layer to avoid over-fitting. The proposed model is trained and tested using different optimization algorithms. The final classification is achieved using a sigmoid layer. We tested the proposed structure on T1 weighted contrast-enhanced magnetic resonance images (T1-w MRI) that are collected from three datasets. A total of 3586 images containing two classes (i.e., bengin, and malignant) were used in our experiments. The proposed model reach highest accuracy 99.8%, and optimal error 0.005 using Adam when compared with other six well-known CNN architectures.

In recent years, indescribable suffering from various kidney diseases has been experienced by people all over the world. The situation has been significantly worse because of chronic kidney disease (CKD). Only through an early diagnosis of CKD may kidney disease be hinder in its early stages from progressing. However, it is easier to detect the chronic kidney disease with the aid of machine learning (ML) classifier algorithms sooner than any other existing methods. The present work proposes an approach for potentially predicting CKD infection while considering patient health dataset information into consideration, employing nine distinct ML algorithms; random forest (RF), naïve bayes (NB), support vector machine (SVM), decision tree (DT), logistic regression (LR), extreme gradient boosting (XGB), adaptive boosting (ADB), k-nearest neighbors (KNN), and neural network (NN). Machine learning algorithms had been utilized and conducted in four experiments, then they were compared using five performance measures; F1-score, precision, accuracy, recall and run time are used to evaluate the performance. Results had shown that AdaBoost (ADA) outperformed other techniques with achieved accuracy of 99.17%.

With the increase in smart devices, performance of traditional networks is limited by this huge amount of generated traffic flows. A scalable and programmable networking solution can be achieved in software defined networks (SDNs) through the separation between the control plane and the data plane. This advantage can allow machine learning (ML) applications to control and automate networks. Concurrently, network slicing (NS) is a promising technology. It is necessary to meet the variety of service needs and requirements. It provides the network as a service (Naas). So, combining NS and ML in SDNs can achieve good network resources management. This paper focuses on applying real-time network traffic analysis to assign each traffic to its suitable network slice according to traffic flows classification. In the proposed model, robust scale is used to scale the features instead of max/min normalization. Also, the k-means clustering algorithm is used to separate the dataset into the optimum number of different clusters (slices). Five different supervised models are applied to achieve high classification accuracy. The highest accuracy that can be obtained from feed-forward artificial neural network is (98.2%), while support vector machine (SVM) with linear function gives an accuracy of (96.7%). The challenges faced are collecting data from SDN’s controller to apply real-time traffic flow classification, which is a primary step to assign each flow to its suitable network slice (Bandwidth)

A crucial area of research is traffic classification, particularly in light of the advancements in machine learning in software-defined networking. Software-defined networks, which divide the control and data planes, can be automated and controlled by machine learning. Because traditional procedures could not keep up with the expanding use of encryption, the use of techniques for this purpose has increased. In this study, 15 features (the quantity of packets communicated, the average transmission time, and the number of instantly transmitted packets) were used to build traffic flows on the SDN for several protocols, including WWW, DNS, FTP, ICMP, P2P, and VOIP. A real-time dataset was produced by gathering data based on the features that were generated over the SDN controller in the physical network. We use the dataset to test and train a variety of machine learning models, including Random Forest, K Nearest Neighbor, Support Vector Machine, Logistic Regression, Decision Tree, and Naive Bayes. With a 99.8% accuracy rate, Decision Tree emerged as the most successful model for the traffic classification challenge. In order to provide the best classification performance with the lowest cost flow features for traffic classification in SDN, this approach has been identified as machine learning