Abstract

Principal Component Analysis (PCA) is a commonly employed technique to help reduce the dimensionality of data to feed machine learning algorithms. This talk will derive the PCA and point out its key assumptions that highlight when this algorithm is appropriate.

“The Karhunen–Loeve (KL) Transform is the most advanced mathematical algorithm available in the year 2008 to achieve both noise filtering and data compression in processing signals of any kind.’’ This quote from C. Maccone highlights the importance of the KL Transform and the talk will explain the KLT . Furthermore, the KLT will be demonstrated by using it to extract a signal buried in noise.

PCA and KL Expansion are terms often used interchangeably. While they are similar, they have important differences. This talk will highlight the context in which both of these are used and point out the similarities and differences between both.