Applying Random Forest to the MNIST handwritten digits dataset

This is a very simple example of using Scikit-learn to train a random forest ensemble on the ubiquitous MNIST handwritten digits dataset.

The dataset is downloaded a set of 70k samples, each of which has 784 elements - this corresponds to a 'flattening' of the original 28x28 image which gives a 'table' of 70k rows and 784 feature columns.