Question 1

What is PCA?

Accepted Answer

Principal Component Analysis (PCA) is a dimensionality reduction technique that transforms data into a new coordinate system by finding eigenvectors of the covariance matrix.

Question 2

How is the covariance matrix computed?

Accepted Answer

First center the data by subtracting the mean of each feature, then compute the matrix product (X_centered^T * X_centered) / (n-1).

Question 3

What does explained variance ratio mean?

Accepted Answer

Each eigenvalue divided by the sum of all eigenvalues gives the proportion of variance explained by that principal component.

🧪 PCA Step-by-Step