The Machine Learning Development Pipe Lane

Inlineor algebra is designed to get rid a sort of compress and remove that exes data that's correlated with each other in various ways. The idea is to reduce out correlated features where possible. For example, om say you are looking at cubes and taking measurements th different and then throwing ellbows in as perimeters on your model. And so fundamentally, what this whole class of techniques is designed to do is find the subspace that really counts.