Kolmogorov-Arnold Network For Classification and Segmentation

During my graduate-level Machine Learning course at K. N. Toosi University of Technology, I worked on a team-based project to implement the Kolmogorov-Arnold Network (KAN) and its extensions, including Fast KAN, Faster KAN, Convolutional KAN, and UKAN. The project focused on evaluating these architectures for classification and segmentation tasks across various datasets, such as MNIST, CIFAR-10, PhysioNet ICH dataset, and ArasCase Capsulorhexis Segmentation dataset, with an emphasis on performance and accuracy. The project was well-received and earned full marks. You can download the full project report here.