Introduction Principal component analysis is one of the most commonly used techniques in statistical modeling and machine learning. In typical applications it serves as a (linear) dimensionality reduction, allowing one to project high dimensional data onto a lower dimensional subspace.