CNN Model for Image Classification on MNIST and Fashion-MNIST Dataset
Shivam S. Kadam, Amol C. Adamuthe, Ashwini Patil
Abstract
Paper presents application of convolutional neural network for image classification problem. MNIST and Fashion-MNIST datasets used to test the performance of CNN model. Paper presents five different architectures with varying convolutional layers, filter size and fully connected layers. Experiments conducted with varying hyper-parameters namely activation function, optimizer, learning rate, dropout rate and batch size. Results show that selection of activation function, optimizer and dropout rate has impact on accuracy of results. All architectures give accuracy more than 99% for MNIST dataset. Fashion-MNIST dataset is complex than MNIST. For Fashion-MNIST dataset architecture 3 gives better results. Review of obtained results and from literature shows that CNN is suitable for image classification for MNIST and Fashion-MNIST dataset.