less than 1 minute read

PCA, vim

PCA (Principle Component Analysis) is a way dimensionality reduction algorithm. The idea is to form a lower dimensional linear subspace(?) and then project our higher dimensional data onto it such that we retain the highest amount/percentage of variance. PCA is good for data compression, makes us take up less memory and speeds up algorithms by reducing computations at the cost of some information. PCA is also handy for reducing dimensions to 2d or 3d so that data can be visualized. PCA can be bad if we try to use it to reduce overfitting since we already have regularization for that. Also we shouldnt prematurely plan to use PCA in our ML project until we come across memory issues or our algorithm being too slow.

If we accidentally press Ctrl + s in vim and it becomes unresponsive, Just press Ctrl + q.