![[Pasted image 20240615095736.png|375]] # Principal Component Analysis (PCA) **Principal Component Analysis (PCA)** is a popular method used to transform high-dimensional data (large amount of values per data point) into a lower-dimensional space. This is performed by projecting the data into new low-dimensional axes called *principal components (PCs)*. - These are also synonymous with eigenvectors Initially The level of variation within any given principal component can also be called its eigenvalue The selection process of these principal components ## Steps for PCA > [!abstract] **Steps for PCA** > 1. *Centering:* Center the dataset around zero by subtracting the mean values across all entries > 2. *Scaling:* aaa > 3. *Covariance Matrix:* aaa > 4. *Eigen Decomposition:* aaa > 5. Sort Eigen Pairs (Descending): > 6. Order and Select Largest: > ### Eigenvalue Decomposition ![[Pasted image 20240615100207.png|525]]