Yazar "Al-jumaili, Saif" seçeneğine göre listele

Listeleniyor 1 - 3 / 3
  • [ X ]
    Öğe
    Covid-19 ultrasound image classification using svm based on kernels deduced from convolutional neural network
    (2021) Al-jumaili, Saif; Duru, Adil Deniz
    Abstract— Millions of people are infected daily with Coronavirus to this day, which increases deaths daily, that has made the virus an epidemic. Based on the current crisis, the availability of tool kits for test plays a significant role in fighting against Covid-19. According to less of availability tools and time consume by using traditional medical tools kit, that provide motivation for researchers to use the advantages of artificial intelligence (AI) techniques. Due to the ability of integrated with medical imaging, AI is very useful for precise diagnosis and classification for different types of diseases. However, in this study, we introduce an idea that combines a set of pre-trained deep learning convolutional neural network models with a supervised machine learning classifier, Supporting Vector Machines (SVM). The dataset used in this study was Lung ultrasound (LUS). To extract features from images, we utilized four types of CNN models namely (Resnet18, Resnet50, GoogleNet, and NASNet-Mobile). Depending on the experimental outcomes, our proposed method show outperform compared to the other latest papers published. Our results achieved based on the four types of evaluation metrics which are Accuracy, Precision, Recall, and F1-Score, where all evaluations achieved exceeded of 99%.
  • [ X ]
    Öğe
    Covid-19 X-ray image classification using SVM based on Local Binary Pattern
    (IEEE, 2021) Al-jumaili, Saif; Al-azzawi, Athar; Duru, Adil Deniz
    Coronavirus usually transmits from the animal to the human, but now, the virus transmission is between persons. Therefore, scientists and researchers are trying to develop several types of machine learning methods to defend against COVID-19. Medical images play a significant role in this time due to they can be used to recognize COVID-19 accurately. However, in this paper, we used X-Ray images, the images undergone to sharpening techniques to increase the results further. The texture techniques named local binary pattern (LBP) have been used in order to extract features. The features obtained were applied to the support vector machine (SVM). The results we achieved were 100% for all performance measurements. Our results were conspicuously superior compared to the state-of-the-art papers published.
  • [ X ]
    Öğe
    Unique MIMO system using Gaussian signals and the advantage of these signals in sensing CSI and multipath fading
    (John Wiley and Sons Ltd, 2025) Alyassri, Salam; Ilyas, Muhammad; Aljumaily, Mustafa S.; Al-jumaili, Saif; Duru, Adil Deniz
    To improve communication network efficiency, researchers must look at all aspects of transmission and the mechanisms that regulate their evolution as a whole. These features include solutions for dealing with the channel's noise and interference. To decrease interference and increase spectrum efficiency, orthogonal frequency division multiple access (OFDMA) systems employ orthogonal signals. While transmitting and receiving signals, noise and numerous feeds can be done in diverse ways. It has become increasingly common to use 256 quadratic modulation (QAM), which is more vulnerable to noise and has a higher bit error rate (BER). BERs in OFDM systems were high when multiple feeds and noise were present, as demonstrated in this article. Starting with the transmission and reception of Gaussian subband signals, an improved system has been designed that includes numerous stages of development. Thus, the need for “orthogonally” of transmitted signals to increase spectrum efficiency has been eliminated, as has the effect of surrounding channels. We have created a header for every frame that has been transmitted. Several transmitters and numerous receivers send these frames in parallel so that the channel state information (CSI) attributes may be evaluated using parallel processing. Using the identical transmission conditions for both OFDM systems and the proposed system, the simulation results reveal a significant reduction in BER values. This results in BER values of fewer than 10−1 when there are two tabs and 10−1 when there are three tabs for multiple feeding in the OFDM system. This corresponds to BER values of 10−11 in a suggested system when there are three tabs. Some improvements have been made to the proposed design to make it distinctive and qualified to be regarded as a multiaccess system in today's contemporary communication infrastructures.