Description

CNN

Announcement

n th
n 
Outline

CNN Algorithm Using Pytorch Assignemnt 5

Lab assignment 5: CNN
Create Your First Convolutional Neural Network in PyTorch Read in CIFAR-10 Dataset Build a CNN with the following architecture:
l 5×5 Convolutional Layer with 16 filters, stride 1 and padding 2.
l ReLU Activation Layer  Batch Normalization Layer
l 2×2 Max Pooling Layer with a stride of 2  3×3 Convolutional Layer with 32 filters, stride 1 and padding 1.
l ReLU Activation Layer  Batch Normalization Layer
l 2×2 Max Pooling Layer with a stride of 2  Fully-conneted layer with 512 output unit  ReLU Activation Layer
l Fully-connected layer with 10 output units
Set up Adam optimizer, with 1e-3 learning rate and betas=(0.9, 0.999) Report the overall accuracy and accuracy of each class Optional: Manipulate parameters to improve the accuracy (epochs, learning rate, number of filters, number of layers ,etc. ) 