PCA Steps
1) Calculate center of data (across all dimensions)
2) Make the center of the data the origin
3) Fit a line to the data that passes through origin (first principle component)
4) Perpindicular line to PC1 is PC2; Line perpindicular to PC1 and PC2 is PC3 etc
5) Rotate plot so PC1 is horizontal (if graphing)
Loading Scores
Proportion of each of the variables effect on a particular principal component
PCA in 1 sentence
A linear combination representation of features to reduce the number of dimensions that a feature space
Scree Plot
Graphical representation of the percentage of variation each PC accounts for.
Formula: SS(distances of PCx)/(n-1)