COM6018 Data Science with Python

Week 10: Classifier Evaluation

Jon Barker

Copyright © 2023–2025 Jon Barker, University of Sheffield. All rights reserved.

In this lab

We will use this lab session to provide further help with Assignment 2.

• Training your own classifier
• Writing your own pipeline functions
• Plotting ROC curves

Lab Materials

This lab is structured a little differently from previous labs.

• The lab instructions are provided as a Jupyter notebook.
• The notebook provides example code.
• You are encouraged to try out the ideas using a Python script.
• There will not be a solutions notebook.

Obtaining the Jupyter Notebook

If you have cloned and pulled the module's GitHub repository, then you should see:

materials/labs/
├── 100_classifier_evaluation.ipynb

Alternatively, you can download the notebook and data via links on Blackboard.

Getting Help

• Feel free to ask questions about the assignment but please note:

  • We can give general advice but we cannot debug your code or reveal parts of the solution.
  • Questions answered for individuals will be noted and added to the FAQ on Blackboard.

• Unfair means and collusion:

  • You may discuss ideas with other students but you must not share code.
  • You must not ask other students for help debugging code, or offer help.
  • If you are very stuck on the assignment then ask myself or one of the demonstrators.