Principal Component Analysis
Principal Component Analysis, often just shortened to PCA, is a statistical technique used in computer vision (for facial recognition and image compression) as well used for finding structure and patterns in higher (and often very high) dimensional data sets.
My first encounter with principal component analysis was with computer vision, with things such as eigenfaces, optical character recognition, optical word recognition, and lossy compression for image data and for video data.
A deep understanding of principal component analysis requires an understanding of some statistical and linear algebra topics. Specifically, it requires knowledge of standard deviation, covariance, eigenvectors and eigenvalues.