Ahed, Umar Bindir <br/><br/>
COMPARISON OF WELL-KNOWN CONVOLUTIONAL  NEURAL NETWORKS IMAGE CLASSIFICATION  MODELS ON THE CIFAR-10 DATASET / 
UMAR BINDIR AHMED; SUPERVISOR: ASSOC. PROF. DR. UMAR ÖZGÜNALP 
 - xi, 62 sheets;  31 cm.  1 CD-ROM 
<br/><br/>Thesis (MSc) - Cyprus International University. Institute of Graduate Studies and Research Electrical and Electronics Engineering  Department 
<br/><br/>Includes bibliography (sheets 60-62) 
<br/><br/> ABSTRACT<br/>In this research, a supervised learning method is used to assess the performance of <br/>well-known deep convolutional neural network models for image classification on the <br/>CIFAR-10 dataset. VGGNet, GoogLeNet, ResNet, and AlexNet are the four models <br/>that were trained and evaluated using a variety of measures, including loss and <br/>accuracy.<br/>ResNet and VGGNet obtained the greatest test accuracy of 78.15% when comparing <br/>the performances of deep convolutional neural network models during training and <br/>testing. AlexNet and GoogleNet came in second and third, with the highest test <br/>accuracy of 77.5% and 74.41%, respectively. These outcomes suggest that the deep <br/>convolutional neural network models in a supervised method are capable of <br/>successfully categorizing CIFAR-10 images with reasonably small loss values, as <br/>shown in Table 4.1.<br/>The findings underscore the significance of meticulous planning and execution of <br/>training and testing methodologies for deep convolutional neural networks in <br/>supervised learning scenarios to efficiently exploit data with annotations. One may <br/>investigate various strategies, including data augmentation, regularization, and <br/>learning rate scheduling, to enhance the effectiveness of the models. <br/>The result of this investigation emphasizes the importance of precise preparation and <br/>implementation of training and testing procedures in deep convolutional neural <br/>networks in supervised learning in order to maximize their capacity for learning. <br/>Further investigation is necessary to tackle the limitations and challenges that occur <br/>with maximizing the use of data with annotations.<br/>The study at issue offers major findings regarding the impact and implications of the <br/>models utilized on the CIFAR-10 dataset. The findings contribute to an understanding of the strengths and weaknesses of these approaches, guiding future research to enhance the performance of the image classification models.<br/>Keywords: Alexnet, CIFAR-10 Dataset, Convolutional Neural Networks, Googlenet, <br/>Image Classification, Resnet, Vggnet 
<br/><br/><label>Subjects--Topical Terms: </label><br/> Neural networks (Computer science) --Dissertations, Academic<br/>Data sets --Dissertations, Academic