Code samples used to present an introductory talk on scikit-learn, for the Data Science Oxford Meetup: http://www.meetup.com/Data-Science-Oxford/events/204629552/
The original source of these code snippets is http://scipy-lectures.github.io/advanced/scikit-learn/
Some rearranging has been done as well as some visualisation code added, to help run through these as a talk.
1 - iris.py: Demonstration of Principal Component Analysis and Support Vector Machines (supervised learning)
2 - lena.py: Demonstration of K-Means Clustering (unsupervised learning)
3 - faces.py: Use of PCA and SVMs for facial recognition
The quickest way to get a working python distribution on any platform to do machine-learning is to download Anaconda: http://continuum.io/downloads
The Spyder IDE from Anaconda was used during the talk, as it is an easy way to get an IPython console next to the code snippet.
The three python scripts have Spyder cells set up in them (the #%% lines define new cells). During the talk the code was run cell-by-cell, using Shift+Enter.