Iraqi Journal for Computers and Informatics

Semi-supervised learning is the class of machine learning that deals with the use of supervised and unsupervised
learning to implement the learning process. Conceptually placed between labelled and unlabeled data. In certain cases, it enables the large numbers of unlabeled data required to be utilized in comparison with usually limited collections of labeled data. In standard classification methods in machine learning, only a labeled collection is used to train the classifier. In addition, labelled instances are difficult to acquire since they necessitate the
assistance of annotators, who serve in an occupation that is identified by their label. A complete audit without a supervisor is fairly easy to do, but nevertheless represents a significant risk to the enterprise, as there have been few chances to safely experiment with it so far. By utilizing a large number of unsupervised inputs along with the supervised inputs, the semi supervised learning solves this issue, to create a good training sample. Since semi-supervised learning requires fewer human effort and allows greater precision, both theoretically or in practice, it is of critical interest.

Currently, Corona-virus disease (COVID-19), one of the most infectious diseases in the 21st century, is a highly
contagious viral infection with a severe impact on global health. It also affected the global economy very badly. This virus threatens human’s life, and it is necessary to design a monitoring device to monitor the patient’s health remotely to avoid the spread of infection to doctors or nurses. In this paper, a wearable device which contains two sensors are used to measure blood oxygenation, body temperature, and heart rate. Then send these readings to the server to analyze them and send warning notifications to the patient’s assistant phone to inform him of whether the oxygen level is low, the heart rate is irregular or the patient's temperature is high to perform a certain procedure to aid the patient.

Pneumonia is a serious chest disease that affects the lungs. This disease has become an important issue that must be taken care of in the field of medicine due to its rapid and intense spread, especially among people who are addicted to smoking. This paper presents an efficient prediction system for detecting pneumonia using nine pre-trained transfer learning models based on deep learning technique (Inception v4, SeNet-154, Xception, PolyNet, ResNet-50, DenseNet-121, DenseNet-169, AlexNet, and SqueezeNet). The dataset in this study consisted of 5856 chest x-rays, which were divided into 5216 for training and 624 for the test. In the training phase, the images were pre-processed by resizing the input images to the same dimensions to reduce complexity and computation. The images are then forwarded to the proposed models (Inception v4, SeNet-154, Xception, PolyNet, ResNet-50, DenseNet-121, DenseNet-169, AlexNet, SqueezeNet) to extract features and classify the images as normal or pneumonia. The results of the proposed models (Inception v4, SeNet-154, Xception, PolyNet, ResNet-50, DenseNet-121 DenseNet-169, AlexNet and SqueezeNet) give accuracies (98.72%, 98.94%, 98.88%, 98.72%, 96.2%, 94.69%, 96.29%, 95.01% and 96.10%) respectively. We found that the SeNet-154 model gave the best result with an accuracy of 98.94% with a validation loss (0.018103). When comparing our results with older studies, it should be noted that the proposed method is superior to other methods.

Robotic manipulator-forward Kinematics involves the assurance of end-effector arrangements from connecting joint boundaries. The traditional mathematical calculation of controller forward -Kinematics is monotonous and tedious. Accordingly, it is important to execute a strategy that precisely performs forward energy while wiping out the disadvantages of the mathematical calculation technique. Versatile Neuro-Fuzzy Inference System (ANFIS) is a computational knowledge strategy that has been effectively executed for expectation purposes in assorted logical orders. This present examination's essential goal was to evaluate the productivity of ANFIS in foreseeing 3-levels of opportunity automated controller Cartesian directions from connecting joint boundaries. A speculative 3-level of opportunity automated controller has been considered in this investigation. Model preparing information has been obtained by mathematical forward kinematics calculation of the controller's end effector arrangements. Nine datasets have been utilized for model preparing, while five datasets have been utilized for model testing or approval. The ANFIS model's precision has been surveyed by figuring the Mean outright Percentage Error (MAPE) between the real and anticipated end-effector Cartesian directions. Because of Mean Absolute Percentage Error (MAPE), the created ANFIS model has forecast correctness’s of 63.35% and 80.07% in foreseeing x-directions and y-organizes, separately. Accordingly, ANFIS can be dependably executed as a commendable substitute for the customary arithmetical calculation method in anticipating controller Cartesian directions. It is suggested that the precision of other computational knowledge methods like Particle Swarm Optimization (PSO) and Support Vector Machines (SVM) be evaluated