COM6018 Data Science with Python

Week 8: Classification with scikit-learn

Jon Barker

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

In this lab

Using scikit-learn to compare different classifiers.

• Preparing a face recognition task
• Performing feature pre-processing and dimensionality reduction
• Building k-NN, SVM, Random Forest and MLP classifiers
• Using GridSearchCV to perform parameter optimisation
• Using pipelines to streamline the evaluation process
• Evaluating using confusion matrices and per-class precision, recall and F1-scores

The Task

We will use scikit-learn to recognise famous people from photographs.

• We will use scikit-learn's built-in dataset, 'Labeled Faces in the Wild'
• Details here, http://vis-www.cs.umass.edu/lfw/
• It contains 13,233 images of 5,749 famous people.
• It is designed for a face verification task, but we will use a subset of it for a classification task.

You will be using the same data in Assignment 2, so this is a good opportunity to get familiar with it.

Example Data

Examples from the 'Labeled Faces in the Wild' dataset.

Example Data

Examples from the 'Labeled Faces in the Wild' dataset.

The Data

dict_keys(['data', 'images', 'target', 'target_names', 'DESCR'])

images - 3D NumPy array storing original colour images (N x height x width)
data - 2D NumPy array storing pre-processed images (N x n_pixels)
target - array of N labels (stored as integers)
target_names - list of the class names

Preprocessing

The data has been pre-processed.

• Images were cropped to the face.
• They were transformed to greyscale.
• They were resized to 125x94 pixels.

Obtaining the Jupyter Notebook

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

materials/labs/
├── 070_classification_with_scikit_learn.ipynb
|-- ... etc

The lab is 070_classification_with_scikit_learn.ipynb. It does not require any additional data files. (The dataset is built into scikit-learn.)

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

Getting Help

• If you are stuck just raise a hand to ask for help.

• Feel free to discuss the lab with your neighbours.

• Re-read the scikit-learn tutorial notes

  • In the Git repo at materials/tutorials/070_Classification_with_Scikit_Learn.ipynb
  • or online at https://uos-com-6018.github.io/COM6018

• Use the scikit-learn API documentation for reference. https://scikit-learn.org/stable/modules/classes.html