Image classification on the CIFAR 10 Dataset using Support Vector Machines (SVMs), Fully Connected Neural Networks and Convolutional Neural Networks (CNNs). The files are organized as follows:
SVMs -- Image Classification on the CIFAR-10 Dataset using Support Vector Machines. Different types of kernels are used including Linear Kernel, Polynomial Kernel and the Radial Basis Function (RBF) Kernel. By using three kernal in SVM the test set accuracy of Linear kernal is 30.600. other Polynomial Kernel and the Radial Basis Function (RBF) Kernel they are not satisfied the accuracy.
convolutional Neural Networks (CNNs): with three fully convolutional layers was observed to produced a best result that is accuracy of 94%.
please see the Report every concept is explained in detailed. !Report is uploaded.