Transforming a dataset into one with fewer columns is more complicated than it might seem, explains Dr. James McCaffrey of Microsoft Research in this full-code, step-by-step machine learning tutorial.

PCA is an important tool for dimensionality reduction in data science and to compute grasp poses for robotic manipulation from point cloud data. PCA can also directly used within a larger machine ...