Dimensionality reduction is an analytical tool, designed to reduce the number of features in our data. By features, we typically mean results from our measurements. The aim of a dimensionality reduction process is to find the most important features (or combinations of features) present in our data and allow us to drill down into these.
In this section, we will focus on principal component analysis (PCA), probably the most popular dimensionality reduction method. First, we will introduce some background of this method, before looking in detail at an example of applying PCA to a multi-variate dataset. Finally, you will be given the chance to try PCA for edge detection on some neutron imaging data.