نظام الرسائل والاطاريح الجامعية/ المكتبة المركزية

صورة غلاف الرسالة/الاطروحة غير متوفرة
مشاهدة
لايوجد ملف
لايوجد ملف
العنوان باللغة العربية
منصة الرسائل والاطاريح: تصنيف مرض كوفيد -19 باستخدام التعبير الجيني وتقنية التعلم العميق - جامعة بابل
العنوان باللغة الانكليزية
Classification of COVID -19 Disease Using Genes Expression and a Deep Learning Technique
اسم الطالب باللغتين
ايمان حميد هادي - Eman Hamid Hadi
اسم المشرف باللغتين
حسين عطية--Hussein A. Lafta
الخلاصة
Millions of people are impacted by the coronavirus illness . It should also be highlighted that the cytokine storm has grown to be a significant factor in the high death rate. However, because of our limited understanding of the host defense mechanism and the emergence of a cytokine storm to combat this viral infection, efforts to create medicines, vaccines, and treatments have been unsuccessful. Therefore, a greater comprehension of the processes causing immunological dysregulation and the emergence of cytokine storms may provide us with insights into the clinical treatment of severe instances. Infection with COVID-19 illness may be influenced by genetic factors . therefore ,identifying the genes that influence this disease can help to increase the good therapy response. The proposed system consists of two main stages: the feature selection and the prediction stage. Feature selection is performed using the feature sequential selection (FSS) method to select a subset of important genes and improve the prediction accuracy of the proposed model. In general, the proposed system implements the FSS method, which recognizes the most advantageous features at each step, then they are entered into the model to ensure the importance of these features and the accuracy that can be obtained. Moreover, this thesis sought to provide a prediction model based on artificial neural network (ANN) and a convolutional neural network (CNN) to identify genes related to patients with COVID-19 disease (DORF6), genes related to people with mild symptoms (WT) or genes related to people without infection (MOCK), the available dataset was used to achieve the goals of the current thesis is: the COVID-19 dataset. The evaluation was made based (accuracy ).The results obtained showed that the performance of the proposed system is effective, as the prediction accuracy of the original and selected genes was compared before and after applying the FSS method. Using all genes, the prediction accuracy obtained was (30%) with ANN and (30%) with CNN, while the prediction accuracy after applying the FSS method was (97%) with ANN and (82%) with CNN with only 198 genes.
الفئة
المجموعة الطبية
الاختصاص باللغة العربية
الاختصاص باللغة الانكليزية
السنة الدراسية
2022
لغة الرسالة/الاطروحة
اللغة الانكليزية
الشهادة
ماجستير
رابط موقع (doi)
Open access
نعم