Principal component analysis at it's best. Works awfully well on most machine learning problems. Often PCA is not just enough by itself, so you run it before yo

Principal component analysis at it's best. Works awfully well on most machine learning problems. Often PCA is not just enough by itself, so you run it before you run a more sophisticated algorithm (say k-Nearest-Neighbors, or Support Vector Machines). It helps in reducing the training time and execution time of those algorithms.