Pycon UK Introductory Tutorial

This tutorial was delivered at PyCon UK 2014. For a more condensed version, please visit Python for ML. For an introduction to neural networks, please visit Neural Networks in a Nutshell.

Installation Notes

This tutorial requires pandas, scikit-learn and IPython with the IPython Notebook. These can be installed with pip by typing the following in terminal:

pip install numpy pandas sklearn ipython

We will be reviewing the materials with the IPython Notebook. You should be able to type

ipython notebook

in your terminal window and see the notebook panel load in your web browser.

Downloading the Tutorial Materials

You can clone the material in this tutorial using git as follows:

git clone git://github.com/savarin/pyconuk-introtutorial.git

Alternatively, there is a link above to download the contents of this repository as a zip file.

Static Viewing

The notebooks can be viewed in a static fashion using the nbviewer site, as per the links in the section below. However, we recommend reviewing them interactively with the IPython Notebook.

Presentation Format

The tutorial will start with data manipulation using pandas - loading data, and cleaning data. We'll then use scikit-learn to make predictions. By the end of the session, we would have worked on the Kaggle Titanic competition from start to finish, through a number of iterations in an increasing order of sophistication. We’ll also have a brief discussion on cross-validation and making visualisations.

• Section 1-0 - First Cut.ipynb
• Section 1-1 - Filling-in Missing Values.ipynb
• Section 1-2 - Creating Dummy Variables.ipynb
• Section 1-3 - Parameter Tuning.ipynb
• Appendix A - Cross-Validation.ipynb
• Appendix B - Visualisation.ipynb

Time-permitting, we would cover the following additional materials.

• Section 1-4 - Building Pipelines.ipynb
• Section 1-5 - Final Checks.ipynb
• Section 2-1 - Support Vector Machines.ipynb
• Section 2-2 - SVM with Parameter Tuning.ipynb

A Kaggle account would be required for the purposes of making submissions and reviewing our performance on the leaderboard.

Credits

Special thanks to amueller, jakevdp, and ogrisel for the excellent materials they've posted.