Description

ANN

Department of Computing
Announcement

n th
n 
Outline
n Introduction of Pytorch 
n ANN Algorithm Using Pytorch Assignemnt 4

Pytorch
n It’s a Python based scientific computing package targeted at two sets of audiences:
A replacement for NumPy to use the power of GPUs Deep learning research platform that provides maximum
flexibility and speed
n Document: https://pytorch.org/get-started/locally/ 
Lab assignment 4: ANN
Create Your First Artificial Neural Network in PyTorch Read in Iris Dataset Split Data into training and test data with proportion 3:1.
n Build an artificial neural network using CrossEntropyLoss
nn.CrossEntropyLoss().(You can try different hidden layers, different activation functions such as ReLu, LeakyReLu, etc.)
n Train the network and draw the training loss and training accuracy plot. Report your best test accuracy on test data.
n 