A05

Download this page as PDF

Assignment 05: Multi-Layer Neural Network on the Iris Dataset Using PyTorch
(07-Apr to 18-Apr)

Objective

The goal of this assignment is to implement a multi-layer neural network (MLP) using PyTorch to classify the Iris dataset.

Question

Design and implement a multi-layer neural network using PyTorch to classify the Iris dataset. Your implementation should follow these steps:

Dataset Preparation

• Load the Iris dataset using sklearn.datasets.load_iris.
• Convert the dataset into PyTorch tensors.
• Split the dataset into training and test sets (e.g., 80% training, 20% testing).
• Normalize the feature values.

Build the Neural Network Model

• Implement an MLP with PyTorch using torch.nn.Module.
• The model should have:
  • An input layer with 4 neurons (one for each feature).
  • At least one hidden layer with ReLU activation.
  • An output layer with 3 neurons (one for each class) and softmax activation.

Train the Model

• Define the loss function (CrossEntropyLoss).
• Choose an optimizer (e.g., Adam or SGD).
• Train the model for a fixed number of epochs (e.g., 100 epochs).
• Track the loss during training.

Evaluate the Model

Compute accuracy on the test set. Generate a confusion matrix to visualize performance.

Theory

• Neuron and it application in Regression/ Classification.

Boilerplate Code

• + How to create a neural network module.
• + How to create a pytorch dataset from tensors.

Evaluation Criterion

This assignment shall be implemented by students and evaluated by the instructor in lab.